# ==========================
# 1. IMPORT LIBRARIES
# ==========================
import pandas as pd
import json
import ast
import gzip
import re
import itertools
from sklearn.feature_extraction.text import CountVectorizer, TfidfVectorizer
from sklearn.decomposition import LatentDirichletAllocation
from sklearn.cluster import KMeans

from nltk.corpus import stopwords
import nltk
nltk.download('stopwords')

import warnings
warnings.filterwarnings('ignore')

[nltk_data] Downloading package stopwords to
[nltk_data]     C:\Users\forca\AppData\Roaming\nltk_data...
[nltk_data]   Package stopwords is already up-to-date!

import plotly.io as pio
pio.renderers.default = "notebook_connected"

import gzip
import ast  # safe for parsing Python-literal dicts
import itertools
import pandas as pd

# ==========================
# 1. LOAD PRODUCT METADATA
# ==========================
META_FILE = "meta_Clothing_Shoes_and_Jewelry.json.gz"

nike_asins = []

with gzip.open(META_FILE, "rt", encoding="utf-8") as f:
    for line in f:
        line = line.strip()
        if not line:
            continue
        try:
            # Parse Python-style dict
            data = ast.literal_eval(line)
            categories = [c.lower() for c in itertools.chain(*data.get("categories", []))]
            if "nike" in categories:
                nike_asins.append(data["asin"])
        except Exception as e:
            print("Skipping line due to parse error:", e)

print(f"Found {len(nike_asins)} Nike ASINs")

# ==========================
# 2. LOAD REVIEW DATA
# ==========================
REV_FILE = "reviews_Clothing_Shoes_and_Jewelry.json.gz"

reviews_list = []

with gzip.open(REV_FILE, "rt", encoding="utf-8") as f:
    for line in f:
        line = line.strip()
        if not line:
            continue
        try:
            data = ast.literal_eval(line)
            if data.get("asin") in nike_asins:
                reviews_list.append({
                    "asin": data.get("asin"),
                    "reviewText": data.get("reviewText"),
                    "overall": data.get("overall"),
                    "reviewerID": data.get("reviewerID"),
                    "reviewTime": data.get("reviewTime")
                })
        except Exception as e:
            print("Skipping line due to parse error:", e)

reviews_df = pd.DataFrame(reviews_list)
print(f"Found {len(reviews_df)} reviews for Nike products")

print(reviews_df.head())

Found 8327 Nike ASINs
Found 21570 reviews for Nike products
         asin                                         reviewText  overall  \
0  B0000V9K32  the colour i received is not blue as shown but...      3.0   
1  B0000V9K32  Very cute and is really practical. Fits better...      5.0   
2  B0000V9K3W  The watch was exactly what i ordered and I got...      4.0   
3  B0000V9K46  This product came promptly and as described, p...      5.0   
4  B0000V9KNM  Why isn't Nike making these anymore?  I love t...      4.0   

       reviewerID   reviewTime  
0   ACT5DY536GISV  07 27, 2009  
1  A3BVWMS9I8OH8U  12 26, 2009  
2   A5RZS69KSJH00  10 24, 2009  
3  A3F8O512N9UNVM  03 26, 2008  
4  A2EAKTCKFRF7A4  11 16, 2011

# ==========================
# 3. TEXT CLEANING
# ==========================
from nltk.stem import WordNetLemmatizer
stop_words = set(stopwords.words('english'))
# Optionally add brand-specific stopwords like 'nike'
domain_stop_words = {'nike','shoe','shoes','product','pair','wear','feet','foot','running','run'}
stop_words = stop_words.union(domain_stop_words)

lemmatizer = WordNetLemmatizer()

def clean_text(text):
    text = str(text).lower()
    # Basic cleaning
    text = re.sub(r'[^a-z\s]', '', text)  # remove punctuation/numbers
    # split and filter
    words = [lemmatizer.lemmatize(w) for w in text.split() 
             if w not in stop_words and len(w) >= 3]
    return ' '.join(words)

reviews_df['cleaned'] = reviews_df['reviewText'].apply(clean_text)

reviews_df

# ==========================
# 4. VECTORIZE TEXT
# ==========================
# Using TF-IDF vectorization
vectorizer = TfidfVectorizer(max_df=0.95, min_df=5)
X = vectorizer.fit_transform(reviews_df['cleaned'])

# ==========================
# 5a. TOPIC MODELING (LDA)
# ==========================
num_topics = 75  # Adjust to your dataset and interpretability
lda = LatentDirichletAllocation(n_components=num_topics, random_state=42)
lda.fit(X)

# Function to display top words per topic
def display_topics(model, feature_names, no_top_words):
    topics = {}
    for topic_idx, topic in enumerate(model.components_):
        top_words = [feature_names[i] for i in topic.argsort()[:-no_top_words - 1:-1]]
        topics[topic_idx] = top_words
        print(f"Topic {topic_idx}: {' '.join(top_words)}")
    return topics

#feature_names = vectorizer.get_feature_names_out()
#topics = display_topics(lda, feature_names, 10)

# ==========================
# 5b. Display Top 4 Words per Topic AND store them
# ==========================
topic_top_words = {}  # dictionary for mapping topic → top words string

for idx, topic in enumerate(lda.components_):
    top_words = [vectorizer.get_feature_names_out()[i] 
                 for i in topic.argsort()[-7:][::-1]]
    topic_top_words[idx] = ", ".join(top_words)  # store as string
    print(f"Topic {idx}: {top_words}")

Topic 0: ['keeping', 'listed', 'barefoot', 'reading', 'modern', 'self', 'paint']
Topic 1: ['woman', 'often', 'men', 'comfortable', 'fit', 'vacation', 'washed']
Topic 2: ['glove', 'class', 'zumba', 'like', 'great', 'fit', 'love']
Topic 3: ['pink', 'color', 'combination', 'bright', 'hubby', 'love', 'green']
Topic 4: ['four', 'field', 'played', 'comfortable', 'originally', 'great', 'get']
Topic 5: ['wash', 'sock', 'reason', 'summer', 'washing', 'color', 'fun']
Topic 6: ['size', 'correct', 'fashionable', 'lighter', 'great', 'tad', 'fit']
Topic 7: ['future', 'thank', 'nephew', 'great', 'happy', 'love', 'asked']
Topic 8: ['flip', 'flop', 'pricey', 'father', 'shes', 'love', 'spike']
Topic 9: ['muy', 'que', 'para', 'los', 'excelente', 'producto', 'con']
Topic 10: ['boyfriend', 'play', 'basketball', 'ball', 'jacket', 'supportive', 'player']
Topic 11: ['enjoy', 'fake', 'worried', 'rip', 'lady', 'internet', 'ship']
Topic 12: ['great', 'fan', 'package', 'thrilled', 'huge', 'spring', 'gift']
Topic 13: ['picked', 'kinda', 'tighter', 'lil', 'stable', 'feel', 'balance']
Topic 14: ['ever', 'best', 'comfortable', 'ive', 'regret', 'considering', 'great']
Topic 15: ['good', 'training', 'cross', 'trainer', 'blister', 'site', 'support']
Topic 16: ['swoosh', 'thumb', 'fell', 'liked', 'cloth', 'shoelace', 'elastic']
Topic 17: ['advertised', 'rating', 'cold', 'chose', 'promptly', 'weather', 'compression']
Topic 18: ['shirt', 'expecting', 'lite', 'shin', 'recomiendo', 'splint', 'bien']
Topic 19: ['room', 'sister', 'backpack', 'school', 'penny', 'flexible', 'love']
Topic 20: ['others', 'boy', 'definately', 'active', 'adorable', 'recommend', 'cousin']
Topic 21: ['confortable', 'alot', 'soon', 'indoor', 'exact', 'love', 'fit']
Topic 22: ['golf', 'slipper', 'comfortable', 'person', 'bulky', 'sporty', 'fit']
Topic 23: ['birthday', 'gift', 'brother', 'earlier', 'opened', 'love', 'got']
Topic 24: ['worked', 'surprised', 'affordable', 'shipment', 'complement', 'cortez', 'well']
Topic 25: ['force', 'air', 'excelent', 'forever', 'combo', 'expected', 'man']
Topic 26: ['fitting', 'work', 'comfortable', 'thong', 'cannot', 'hour', 'forward']
Topic 27: ['insole', 'like', 'comfortable', 'toe', 'name', 'support', 'ive']
Topic 28: ['pocket', 'item', 'length', 'continue', 'authentic', 'space', 'loose']
Topic 29: ['okay', 'decent', 'besides', 'stretch', 'multiple', 'good', 'great']
Topic 30: ['present', 'particular', 'christmas', 'unfortunately', 'tshirt', 'red', 'boot']
Topic 31: ['rub', 'teenage', 'actual', 'vibrant', 'feeling', 'unique', 'styling']
Topic 32: ['shox', 'wider', 'wait', 'town', 'guy', 'careful', 'cant']
Topic 33: ['wonderful', 'spent', 'great', 'anymore', 'store', 'fit', 'comfortable']
Topic 34: ['marathon', 'treadmill', 'comfortable', 'exercise', 'promised', 'typically', 'sad']
Topic 35: ['recommend', 'anyone', 'great', 'delivery', 'job', 'highly', 'would']
Topic 36: ['grandson', 'thanks', 'he', 'bra', 'love', 'wallet', 'excited']
Topic 37: ['watch', 'band', 'battery', 'rate', 'one', 'time', 'heart']
Topic 38: ['width', 'colour', 'brought', 'wide', 'fit', 'size', 'great']
Topic 39: ['sunglass', 'lens', 'five', 'shopping', 'worst', 'case', 'star']
Topic 40: ['foam', 'yes', 'comfortably', 'website', 'memory', 'told', 'shower']
Topic 41: ['vomero', 'third', 'outlet', 'zoom', 'neutral', 'free', 'waste']
Topic 42: ['awesome', 'comfortable', 'jean', 'great', 'love', 'super', 'look']
Topic 43: ['beat', 'ended', 'uncomfortable', 'fair', 'dart', 'scheme', 'price']
Topic 44: ['apart', 'early', 'enjoyed', 'season', 'state', 'falling', 'sell']
Topic 45: ['held', 'everyday', 'difference', 'growing', 'impressed', 'shade', 'well']
Topic 46: ['breaking', 'cause', 'sock', 'fit', 'little', 'good', 'adjustable']
Topic 47: ['option', 'accurate', 'near', 'navy', 'adjust', 'seemed', 'generally']
Topic 48: ['bag', 'gym', 'clothes', 'hold', 'stuff', 'carry', 'perfect']
Topic 49: ['excellent', 'holding', 'football', 'wife', 'practice', 'good', 'fit']
Topic 50: ['value', 'orange', 'good', 'bright', 'per', 'jersey', 'great']
Topic 51: ['outfit', 'satisfied', 'totally', 'product', 'match', 'recomend', 'quality']
Topic 52: ['max', 'air', 'general', 'walking', 'basically', 'love', 'couple']
Topic 53: ['dirty', 'waiting', 'designed', 'white', 'sooner', 'mention', 'screen']
Topic 54: ['mom', 'cheaper', 'asking', 'breathe', 'loving', 'transaction', 'ahead']
Topic 55: ['team', 'daughter', 'dark', 'christmas', 'service', 'love', 'good']
Topic 56: ['watch', 'glass', 'face', 'hopefully', 'resistant', 'great', 'digital']
Topic 57: ['kid', 'packaging', 'somewhat', 'retro', 'faster', 'fitted', 'womens']
Topic 58: ['reasonable', 'stated', 'complain', 'price', 'terrible', 'believe', 'teenager']
Topic 59: ['ask', 'refund', 'xmas', 'sent', 'attention', 'get', 'received']
Topic 60: ['size', 'small', 'half', 'bigger', 'larger', 'fit', 'order']
Topic 61: ['cleat', 'happier', 'plan', 'wet', 'great', 'trail', 'pick']
Topic 62: ['amazing', 'find', 'anywhere', 'wow', 'glad', 'love', 'couldnt']
Topic 63: ['deal', 'kick', 'choose', 'real', 'dog', 'brown', 'good']
Topic 64: ['gotten', 'many', 'strong', 'shown', 'compliment', 'surprise', 'love']
Topic 65: ['watch', 'wrist', 'easy', 'button', 'activity', 'band', 'feature']
Topic 66: ['nice', 'please', 'customer', 'service', 'exchange', 'size', 'picky']
Topic 67: ['love', 'fit', 'exactly', 'perfect', 'great', 'comfortable', 'described']
Topic 68: ['purple', 'leg', 'fresh', 'sock', 'hate', 'great', 'comfortable']
Topic 69: ['photo', 'pegasus', 'date', 'late', 'poor', 'positive', 'narrower']
Topic 70: ['delivered', 'family', 'durability', 'waterproof', 'period', 'opinion', 'base']
Topic 71: ['jogging', 'slide', 'pic', 'uniform', 'comfortable', 'great', 'helped']
Topic 72: ['experience', 'wonderfully', 'ipod', 'recommendation', 'aware', 'tread', 'specifically']
Topic 73: ['sock', 'dry', 'warm', 'padding', 'strap', 'cut', 'boot']
Topic 74: ['pant', 'fit', 'smooth', 'great', 'love', 'ride', 'good']

# ==========================
# 5c. ASSIGN TOPICS TO REVIEWS
# ==========================
topic_assignments = lda.transform(X)
reviews_df['topic'] = topic_assignments.argmax(axis=1)

# Map the dictionary into the DataFrame
reviews_df["topic_top_words"] = reviews_df["topic"].map(topic_top_words)

reviews_df[["reviewText", "topic", "topic_top_words"]].head()

# ==========================
# 6. SAMPLE INSPECTION
# ==========================
# Print a few reviews per topic with ASIN
for t in range(num_topics):
    print(f"\n=== Topic {t} ===")
    sample = reviews_df[reviews_df['topic'] == t].head(5)
    for _, row in sample.iterrows():
        print(f"ASIN: {row['asin']}, Review: {row['reviewText'][:100]}...")

=== Topic 0 ===
ASIN: B00012O2O2, Review: This is a great little watch for the price. I wear it to work and get lots of compliments. I am temp...
ASIN: B000E8F2ZG, Review: This is a very nice and modern fashioned clock.The size is as espected. Highly recommended....
ASIN: B000G42Z2Q, Review: IT'S OK,...
ASIN: B000J2F8GK, Review: TheyAreCofterbleAndNiceGoodGripAnd lookGreatKevenGarrantIs awesomeBecause he madeAwesomeShoe,,...
ASIN: B000QU6VL6, Review: This is a great watch.  I've never seen a design quite like it; it's quite an eye-catcher.  It's als...

=== Topic 1 ===
ASIN: B0007RADZ8, Review: I FOUND THESE AND ORDERED THEM IN 2 DIFFERENT COLORS. MY SON LOVED THEM&#8230;THE PRICE WAS GREAT. H...
ASIN: B0007RADZ8, Review: It's tough to find a great running shoe.  I've been told throughout life that asics, new balance, an...
ASIN: B0007RADZ8, Review: My son picked these out because of the color and style and he prefers Nike.  Since he has wide feet,...
ASIN: B0007TQLC0, Review: I love the watch, but had to return it, as the band was tiny, not just small. I have a 5 1/2" wrist,...
ASIN: B000AI2NJI, Review: My daughter used this backpack 2 years in a row for 8th and 9th grade. She could have used it anothe...

=== Topic 2 ===
ASIN: B0007RADZ8, Review: good colors and good design, a very fair price, and i feel the shoes very confortables for jogging a...
ASIN: B000TYS41E, Review: This is my 3rd one. It has all the features you need and conforms to your wrist like a glove....
ASIN: B000UCX0BO, Review: started to tear along the rim above the laces but not too bad and dont really mind it. shoes are com...
ASIN: B0013UXIAK, Review: The shoe was brand new as posted, it fit well, and I am happy with my purchase. I would recommend th...
ASIN: B0013UXIAK, Review: Wear a 7 1/2 got an 8 the shoe still felt too small but after a couple of daythe began to feel comfo...

=== Topic 3 ===
ASIN: B00019FWE4, Review: Nike Oregon Series Digital Super Watch - Black/Black - WA0024-001I purchased watch as a gift and my ...
ASIN: B00019FWE4, Review: This is the second watch like this I have purchased and both are awesome.  The watch is very comfort...
ASIN: B0007RADZ8, Review: Gr8 fit, and feel. gr8 color combinations a very style strong shoe enjoying them. Would like to see ...
ASIN: B000NKK5H0, Review: After using them for only a couple of weeks, I reached into my front coat pocket to find them broken...
ASIN: B000SBW9OG, Review: Ordered as a gift for my brother in law and it fits him perfectly.He has been wearing them for a few...

=== Topic 4 ===
ASIN: B0006NGUE6, Review: Just about right for the price and actually it is exaclty what I was looking for, fits perfectly, gr...
ASIN: B0007IA7O4, Review: Like many things I shop for on Amazon, I was looking for a pair of basketball shoes that were a good...
ASIN: B0007RADZ8, Review: I really like Nike shoes and clothing. I'm comfortable with the shoes. It is good stuff to buy. I lo...
ASIN: B0007RADZ8, Review: I heard reviews that Nike Revolution was not a commercial success. But when I saw this and read othe...
ASIN: B0007RADZ8, Review: I bought these shoes originally for just wearing to the gym and playing tennis in.  But these babies...

=== Topic 5 ===
ASIN: B0007SMKOO, Review: I liked it a lot! I live in Brazil and here cost +- 150,00 dollar..very expensive!...
ASIN: B000A8BU6A, Review: watches were very nice. our clients loved them. boss loved them. nicely packaged....
ASIN: B000A8BUEC, Review: Ok ... for starters, I paid $38.00 for this watch ... this watch locally is very hard to even find &...
ASIN: B000E8H0O2, Review: Moderately priced, stylish, attention to detail, not overwhelming, easy to operate.  Would have been...
ASIN: B000G42Z2Q, Review: Very light. Feel great and look nice as well. I only wish they came in a color that had black instea...

=== Topic 6 ===
ASIN: B00012O2S8, Review: i have owned one of these nike presto before and i have been thinking of buying a new one. i love th...
ASIN: B00019FWE4, Review: I had this watch for exactly 12 DAYS, exactly. I liked the features and the size and weight etc. It ...
ASIN: B0007RADZ8, Review: I ordered a 9.5 but when I tried them on they felt like a size ten I lay awake in regret wonder what...
ASIN: B0007RADZ8, Review: Good running shoes to consider. It's light and fits well into the feet if you order the correct size...
ASIN: B0007RADZ8, Review: Nice shoe, nice fit, Looks like a tennis ball got murdered to make the laces, so they'll see you com...

=== Topic 7 ===
ASIN: B0006NGUE6, Review: I find these shoes superior to another pair of Nike' golf shoes I previously owned.  They fit well a...
ASIN: B0007RADZ8, Review: I did measure my foot and it was 27 cmi orderd consequently in USA size but it should have been 1/2 ...
ASIN: B000A8BU1U, Review: Thank you very much.  Great price and fast shipping.  Just what I needed....
ASIN: B000AORSVU, Review: Durable. Excellent quality. These were the 2nd pair as a gift for our son. He absolutely loves them....
ASIN: B000AORSVU, Review: Received these sunglasses very promptly.  Everything was as advertised.  Great SELLER!!!  Would defi...

=== Topic 8 ===
ASIN: B0006NGUE6, Review: Good quality shoe that fit as expected. Shoe was purchased for my father and he loved the look and s...
ASIN: B00073G928, Review: Bought as a gift for my running daughter.  Directions are challenging and she's not thrilled with th...
ASIN: B0007RADZ8, Review: Great shoe .... Better than what I expected! Great fit and also very comfortable! I would recommend ...
ASIN: B000E93J64, Review: I bought this watch for my daughter who is on college. She has very small wrists and after trying ot...
ASIN: B000E95LQK, Review: A light, very functional watch for both runners and fashion wear. I always get compliments on it's l...

=== Topic 9 ===
ASIN: B0006NGUE6, Review: Excelente producto, muy bonitos y comodos, los acabados y la presentaci&oacute;n son excelentes, los...
ASIN: B0007QZ0AW, Review: they was right on the money with this one, everything was what they said about the shoes  in size, e...
ASIN: B0007RADZ8, Review: very good, I am very satisfied with my purchase, recommended. compren sin duda.compren sin duda.comp...
ASIN: B000A8BUCO, Review: Es un excelente monitor cardiaco pero al entrar al agua esto cambia completamente, a diferencia del ...
ASIN: B000BJVY5K, Review: muy buena la presentaciony la talla pedida si inconvenietes un buen envio tal como lo que vi en la f...

=== Topic 10 ===
ASIN: B0007RADZ8, Review: i can train for hours and the shoes are still coolone of the best running shoes i had compare to all...
ASIN: B0007SMKOO, Review: Great for playing sports or working out. Its small, it does its job, and comfortable....
ASIN: B000E8GZYI, Review: Cute, but you'd have to have the tiniest wrist in the world for it to fit comfortabtly...
ASIN: B000E8H0O2, Review: only thing i don't like is the volume of the alarm, too low....
ASIN: B000EHIVAK, Review: It was very difficult to find a retailer that offered youth Nike watches.  I appreciate the analog d...

=== Topic 11 ===
ASIN: B0006NGUE6, Review: These shoes turned out to be a very pleasant surprise.  Excellent fit and comfort with everything yo...
ASIN: B0007IA7O4, Review: They copied off of Kobe's  shoes. The nike zoom atteroes on the other hand it has really good ankle ...
ASIN: B0007RADZ8, Review: I love hess shoes. I do Taebo and run daily. These shoes are really light weight and can take a beat...
ASIN: B000P6C3E0, Review: The shoe was in brand new condition and fit perfectly! I bought it as a gift and saved me lots of mo...
ASIN: B000V4YZ1K, Review: They seem really small. And I don't have big feet. They aren't small (ladies) but not huge either. J...

=== Topic 12 ===
ASIN: B0000V9KRS, Review: I totally love this watch. It is much nicer and more beautiful than I thought when I bought it for a...
ASIN: B0007RADZ8, Review: We were hoping to get a decent pair of NIKE's for the kid at a good price. These are not them. They ...
ASIN: B0007RADZ8, Review: Purchased as a gift.  Shoe fit very comfortably.  Recipient LOVED them.  Found new faith in buying s...
ASIN: B000A8BTOI, Review: I loved the look of this watch that was on Amazon.com and ordered it.  When I received it, the watch...
ASIN: B000E8H0LA, Review: I bought this watch as a birthday gift for my husband and he loves it!  It's beautiful, made well, a...

=== Topic 13 ===
ASIN: B0000V9KRI, Review: I mean, Roxy rocks, but I'm kinda dissapointed with the material. The purse lokks a lil' bit cheap....
ASIN: B000276C7Q, Review: I Love this watch.  My only problem is it's fit.  I have very small wrists.  I wish there was a way ...
ASIN: B0007RADZ8, Review: These shoes are great for exercising even running if that suits you. Great support, durable and my o...
ASIN: B0007RADZ8, Review: Full disclosure is that I didn't purchase this shoe from Amazon, but I love it so much I wanted to l...
ASIN: B0007RADZ8, Review: Comfortable, but could be wider through the toes. My big toe on each foot occasionally goes over the...

=== Topic 14 ===
ASIN: B0006NGUE6, Review: I was looking at more expensive shoes but thought I would give these a try. Great for a second pair ...
ASIN: B0006NGUE6, Review: Best shoes I have ever owned.  They look amazing!! And, they feel and fit perfectly.  I would buy ag...
ASIN: B0007QZ0AW, Review: This is the best shoe I have ever had. It's great looking, and wears like iron. Like you I have had ...
ASIN: B0007RADZ8, Review: My teenage son is a fan of Nike shoes.  I myself do not buy Nike footwear but he thinks they are ver...
ASIN: B0007RADZ8, Review: Can't really complain about these shoes - they were relatively cheap, comfortable and semi-stylish.....

=== Topic 15 ===
ASIN: B00012O2MO, Review: My best friend has this watch and uses it for her work in the hospital. I loved it, so I got it and ...
ASIN: B0001YMVTS, Review: The product never arrived. The vendor told me it was US postal service issue. The US postal service ...
ASIN: B0007QZ0AW, Review: I purchased the Nike Womens VXT Cross Trainer Shoes, which are identical, with the exception of the ...
ASIN: B0007RADZ8, Review: good product...
ASIN: B0007RADZ8, Review: Nike Men's NIKE REVOLUTION 2 RUNNING SHOES 11 Men US (BLACK/WHITE/VARSITY). I use them to go to the ...

=== Topic 16 ===
ASIN: B0006NGUE6, Review: This product was exactly what I expected. The fit is comfortable and I would recommend this product ...
ASIN: B0006NGUE6, Review: The shoes look great and the process was smooth and eazy. I recommend the buyer look closely at the ...
ASIN: B0007RADZ8, Review: These felt a bit narrow and I ultimately ended swapping them out for the wider version.The shoes are...
ASIN: B000A8BUDI, Review: So far I like the watch. It is a little big, but the features are easy to access and set. Not so sur...
ASIN: B000G42Z2Q, Review: i like the product's material, it's so goodit's somewhat small for my right sidei dislike the shoela...

=== Topic 17 ===
ASIN: B0000V9K46, Review: This product came promptly and as described, pleasure doing business with them!-d...
ASIN: B0006NGUE6, Review: Great fit your just have to buy a 1/2 size higher. The shoe very comfortable and grips well....its t...
ASIN: B000SBW9OG, Review: sturdy and handy sandal for summer wear and around the pool or in the yard. I have had these for sev...
ASIN: B0018E63MG, Review: The setup process is a little bit complicated.  But after that, it works very well....
ASIN: B001G20EMA, Review: i did buy this shoes to go mountain bike riding love the shoes this is the type of shoes i like for ...

=== Topic 18 ===
ASIN: B000F9LFBE, Review: Nice watch. Elegant and comfortable. If you are a Arsenal fan, you have to get it, not kidding, just...
ASIN: B000V4YZ1K, Review: very nice :D :D :D A+++++++++++++++++++++++++ lo mejor bien :3 d  d d fdd df fd dc d df f...
ASIN: B0013UXIAK, Review: I really like, I were spectacular, I can walk comfortable way, I fulfilled my expectations, thank yo...
ASIN: B001718DSM, Review: Excelente la compra llego tal y cual como se esperaba lo recomiendo para futuras compras llego a la ...
ASIN: B0018VL7TS, Review: The quality is so good on this shirt. My son loves it! It really is just as I expected and I am happ...

=== Topic 19 ===
ASIN: B0000V9K32, Review: Very cute and is really practical. Fits better on smaller wrists which is my case. I wear them every...
ASIN: B0007RADZ8, Review: The fit and overall feel of these shoes are fantastic !! I would buy they again and recommend them t...
ASIN: B0007RADZ8, Review: Nice confort , perfect size and incredible soft to run.the timing was pefect and the model is veru b...
ASIN: B0007RADZ8, Review: Need to get them replace as the shoes run smaller than the size I typically use. Please get back to ...
ASIN: B000A8BTO8, Review: My son, who is in middle school, loves this watch. The angle makes it easy to see, and it is pretty ...

=== Topic 20 ===
ASIN: B0006NGUE6, Review: Purchased this as a Christmas gift for my cousin as he just started picking up golf as a sport. The ...
ASIN: B000V6W0IS, Review: it looks good, i think it is not a leather made(It is the reason i bought it), but still OK!...
ASIN: B0010WOV8O, Review: Gift for my boyfriend. He like them. The rounding on the front was an intial concern, but he liked t...
ASIN: B0013UXIAK, Review: he product is of excellent quality, and the time when the delivery has been made &#8203;&#8203;very ...
ASIN: B0013UXIAK, Review: Bought these for the style but the performance is what makes me give it 5 stars. Great hoops shoes, ...

=== Topic 21 ===
ASIN: B00012O2S8, Review: This is a great watch.  It is comfortable and waterproof.  I wore out the first one and purchased th...
ASIN: B0007RADZ8, Review: Great shoes, received in excellent condition. Got a .5 size larger for a better fit. They fit fantas...
ASIN: B0007RADZ8, Review: Its a pretty good shoe, to buy if you are looking for a pair of shoes to wear, if you are looking fo...
ASIN: B0007RADZ8, Review: I love e The colors are very nice. Best quality. Nike is confortable sneakers and is perfect for run...
ASIN: B0007RADZ8, Review: Good Shoes, meet my expectations, color is right, they are confortable and can be worn at the gym an...

=== Topic 22 ===
ASIN: B0001YMVVQ, Review: It's a great watch. I bought a Nike watch about a month earlier, loved it and bought this for a frie...
ASIN: B0006NGUE6, Review: GREAT GOLF SHOES LOVE THEM A A A A A A A A A A A  A A A A+...
ASIN: B0006NGUE6, Review: Fast Shipping and was excellent deal.  So far I have only worn it once, and it felt great.  Didn't h...
ASIN: B0006NGUE6, Review: Great shoes and my grandson loves them and will help him with his golf.  He now no longer uses my sh...
ASIN: B0006NGUE6, Review: I think it is a great golf shoe. comfortable and very functional, it does it's job. Hower, I am glad...

=== Topic 23 ===
ASIN: B0006NGUE6, Review: This item was a birthday gift for my husband. He loves them! He said they were about the most comfor...
ASIN: B0006NGUE6, Review: I purchased these as a surprise for my boyfriend and he loves them. Says they are very comfy and hav...
ASIN: B0007QZ0AW, Review: Shoes arrived clean and laced up, ready to go!  They were in excellent condition from the box.  They...
ASIN: B0007RADZ8, Review: Its a nice shoes feet exactly.. That was gift of  my brother.. I love the brand NIke...The other one...
ASIN: B000E8H01K, Review: Great watch, great price.  This is an awesome watch for weekend wear and sporting activities.  It is...

=== Topic 24 ===
ASIN: B0002764JW, Review: Recieved in good shape and when expected.  Just that the mens size was smaller than I expected.  But...
ASIN: B0006NGUE6, Review: Very please with my recent purchase of these Nike Golf Shoes.  So far they seem to be holding up wel...
ASIN: B0007RADZ8, Review: I wore them within ten minutes of delivery.  Normally I don't do that, but they looked &#34;cool&#34...
ASIN: B0007RADZ8, Review: Nice styling. Like the bright lacings. Seem to be fine when running .Good value for the money. Easy ...
ASIN: B0007RADZ8, Review: I think my brother gonna like it. So happy that I ordered it, and it is so affordable, I will recomm...

=== Topic 25 ===
ASIN: B0006TZU7S, Review: This watch is awesome but I barely got it and the display screen fell off!!! makes me mad....
ASIN: B0007RADZ8, Review: what I expected...
ASIN: B0007RADZ8, Review: Just what I expected!...
ASIN: B000BWLOH0, Review: Love this watch.  Love this team.  Comfortable, ergonomically correct, and well mad and durable.  Th...
ASIN: B000E8J5R2, Review: It is a good product of very good quality, very resistant and especially the mark  sit down satisfie...

=== Topic 26 ===
ASIN: B00012O2MO, Review: It's cute, works reasonably well, but won't last long.  Really cheap guts on this one.  That's too b...
ASIN: B0007RADZ8, Review: I'm a valet attendant, and since I'm running all the time, I go through running shoes pretty quickly...
ASIN: B0007RADZ8, Review: The color is great, which is what attracted me to these shoes. Comfy & Light.  Nice fit, as always N...
ASIN: B0007RADZ8, Review: Love the fit and the shoes R all that was expected. Never had so many compliments. Look forward to m...
ASIN: B0007RADZ8, Review: I purchased this as a gift for Christmas so I cannot comment on the shoe as of yet.  But I really li...

=== Topic 27 ===
ASIN: B0006NGUE6, Review: comfortable low profile shoes that are great for holing you to the ground. Like the feel and certain...
ASIN: B0006NGUE6, Review: Made out of cheap material in an effort to keep weight/cost down. No natural place for the upper to ...
ASIN: B0007RADZ8, Review: These shoe gives good support for short distance running. I usually use this when I do less than 5 m...
ASIN: B0007RADZ8, Review: The Nike Revolution 2 is a good shoe for light running days (1-3 miles). I have been training in thi...
ASIN: B0007RADZ8, Review: These shoe gives good support for short distance running. I usually use this when I do less than 5 m...

=== Topic 28 ===
ASIN: B0007RADZ8, Review: Great fit for my feet. Wish they were a little higher but I expect that with running shoes.Love the ...
ASIN: B000E8J4ZA, Review: I LOVE this watch!  I am going to be purchasing more of the same watches in other colors.  This has ...
ASIN: B000FAG86K, Review: Could have a longer strap but its very good. Durable and can hold a good amount of items in it. I us...
ASIN: B000FAK320, Review: My previous bag fell apart so I purchased this one.  Nice construction, roomy whithout being too big...
ASIN: B000G163DG, Review: Have not tried switching the lenses out -- but love the design and fit -- not too tight in the templ...

=== Topic 29 ===
ASIN: B0006NGUE6, Review: Just started playing golf, and I'm getting a little more serious about, so I wanted to pick up a pai...
ASIN: B0007RADZ8, Review: Decent running shoes...
ASIN: B000A8BU4M, Review: I wear it daily.  the buttons depress accidentally though so it does not stay on the "time" screen....
ASIN: B000E8F2PG, Review: The plastic on the band is a bit harder than I expected.  I thought it would be a softer feel... Ins...
ASIN: B000EHP4XM, Review: Bought this as a gift for my preteen daughter.  She loves it.  It is almost too small though, it wou...

=== Topic 30 ===
ASIN: B00012O2S8, Review: First I have bought these watches in black colour for myself. I loved them! Then I decided to buy an...
ASIN: B0007RADZ8, Review: The fit is excellent.  The style is great.  It is just the shoe for my workouts in the neighborhood....
ASIN: B0007RADZ8, Review: They were a present for my brother, who worn them and found them a little small, thus 4 stars and no...
ASIN: B0007RADZ8, Review: Hi thereSize small, I bought another shoe brand Nike.I hope before apologeticThank you very much. . ...
ASIN: B000A8BU8S, Review: Purchased this as a present for my teenage daughter and she loves it. She never wore a watch before ...

=== Topic 31 ===
ASIN: B0006NGUE6, Review: Very nice look & feel to the shoe. Soft comfortable sole. Now if only my swing would feel as comfort...
ASIN: B0006NGUE6, Review: Very satisfied with the performance of the shoe in the field. Great fit, stability, and support when...
ASIN: B0007RADZ8, Review: IT DOES NOT FEEL GREAT RUNNING COMPARES TO SHOES WITH SHOCKS . IT FELT FLAT LIJE WEARING SANDALS. BO...
ASIN: B0009S1DMC, Review: I have the black w/ grey polarised lenses version, and I cannot be more pleased.  As a runner, I app...
ASIN: B000A8BTPM, Review: This is a good deal for the price I paid on Amazon. Even with $8.00 in shipping I still managed to s...

=== Topic 32 ===
ASIN: B000AI2QTK, Review: So GREAT.this was my first time usig amazon and this guy was so exelent.ThanksAntonio...
ASIN: B000V4YZ1K, Review: Overall length is appropriate for my foot, but strap was way too tight. I have very average, medium ...
ASIN: B000V6W0IS, Review: No chance to try them because they are too big and i did not wear them at all but the y look nice...
ASIN: B001GULR14, Review: I returned them because I didn't like the way they looked on me. The company however was great with ...
ASIN: B001IDPYNG, Review: I purchased these shoes for running. They are fablious! I will buy Nike shox again! They are sized a...

=== Topic 33 ===
ASIN: B00012O2RO, Review: I've found this watch very nice for a really good price (I needed it fast, so shipment was almost 50...
ASIN: B0006NGUE6, Review: The shoes fit perfect. I love them glad I didnt pay attention to the comments about them being to na...
ASIN: B0006NGUE6, Review: Very nice shoes.  Might want to size up .5 your normal size.  Plenty of comments about them on the l...
ASIN: B0007RADZ8, Review: Looked better than i expected it fit wonderful and very cumffy u should buy best shoe good price why...
ASIN: B0007RADZ8, Review: Already owned a pair so when they were getting worn out, one quick stop to Amazon and two days later...

=== Topic 34 ===
ASIN: B0006NGUE6, Review: Great shoes for an excellent price.  The fit was perfect and the swoop was a bit darker than expecte...
ASIN: B0007NDSA4, Review: amazing fit and it is so comfortable! didn't try this on before so i chose my side and it is true to...
ASIN: B0007RADZ8, Review: Received these shoes and instantly fell in love.  They fit perfect and look amazing!!!  I've gotten ...
ASIN: B0007RADZ8, Review: Very nice shoes...my son loved the color and comfort and the lite weight. Would recommend this shoe ...
ASIN: B0007RADZ8, Review: This shoe is inaccurately sized at 11 1/2 - too small. I should have returned them, but thought they...

=== Topic 35 ===
ASIN: B0006NGUE6, Review: I have had these shoes for 3 months now and am really pleased with them!!! They are so comfortable a...
ASIN: B0007RADZ8, Review: I really like the Nike Revolution 2 shoes. I run an average of 3 miles a day and have no complaints ...
ASIN: B0007RADZ8, Review: I gave these shoes 5 stars because I really like the way they look and feel, very comfortable and ve...
ASIN: B0007RADZ8, Review: Just what I needed. A quality shoe, good looking, well built & reasonably priced. the packaging was ...
ASIN: B0007RADZ8, Review: this product is great nice design of the shoe and fits just right good for running and also for walk...

=== Topic 36 ===
ASIN: B0001YMVSY, Review: I bought this watch for my boyfriend for Christmas. He absolutely loves it and wears it all the time...
ASIN: B0002764JW, Review: Great look, fit very well , you won't remember you have it until you need it , unisex, my wife and m...
ASIN: B0007RADZ8, Review: Awesome product , thanks...
ASIN: B000A8BTO8, Review: My grandson loved the Nike watch I gave him....it was a good purchase and the "in" thing for a sport...
ASIN: B000E8H0BU, Review: I purchased this as a gift for my grandson.  He enjoys it very much....

=== Topic 37 ===
ASIN: B0000V9KRS, Review: when it arrived, it had a dead battery--SO, I had to invest in a new battery before I could guarante...
ASIN: B00012O2MO, Review: I have used a similar watch (Arctic Blue) for several years.  I love it.  When I saw the same style ...
ASIN: B00012O2S8, Review: About the price and product itself I have no complaints - it is nice, different, stylish, but it doe...
ASIN: B00019FWE4, Review: I've had this watch for about a year and a half, and for a long time, I really loved it. The best th...
ASIN: B0001YMVTS, Review: i had some problems with this order, the bill didn't arrive with the watch to the p.o. box and it co...

=== Topic 38 ===
ASIN: B0006NGUE6, Review: My husband liked the colors of these shoes, but the width was wrong. Did not have this shoe in a 7 w...
ASIN: B0006NGUE6, Review: The shoes look good and comfortable. I have worn them for about half a dozen times now. The size is ...
ASIN: B0007RADZ8, Review: These shoes are pretty sturdy and provide enough support for me (a heavy-ish runner).  Nikes I buy t...
ASIN: B0007RADZ8, Review: I love it, it is cool and comfortable. and the size fit as expected. and i claim the mail quickly....
ASIN: B0007RADZ8, Review: My son is hard to fit with shoes usually being too tight in the width or not fitting just right in t...

=== Topic 39 ===
ASIN: B0006NGUE6, Review: Excellent Shoes, very comfortable and nice. This case is wide and smooth when you are walking with t...
ASIN: B0007RADZ8, Review: I've only taken these out running about 3 times so far, but I've liked them. They are comfortable an...
ASIN: B0009RM2FU, Review: I forgot that I purchased these through Amazon. They lasted about 5 years of rugged use (beach volle...
ASIN: B000A8BTO8, Review: wow!!! what a great watch.  At first when I got it I thought it was a little hard to understand, but...
ASIN: B000A8BTQ6, Review: I bought the watch on monday and got it on Saturday. Very good delivery!...

=== Topic 40 ===
ASIN: B0007RADZ8, Review: Awesome just what I ordered nike always has the best shoes Adidas suck because the are not nike qual...
ASIN: B000E8CS30, Review: Product.not reliable.battery went flat after 6 months. Replaced new one went flat again in 2 weeks. ...
ASIN: B000P6GKOO, Review: On the website nike.comi can get it for $70 when its not sold out.if it is i can get it for $110 usi...
ASIN: B000SBW9OG, Review: These are great shoes.  I use them for everything during the hot Texas summers. I highly recommend t...
ASIN: B000V4YZ1K, Review: Soft as hell! Thought that these wouldn't be as great as the Adidas slippers I owned but these where...

=== Topic 41 ===
ASIN: B00012O2RO, Review: This is my third Nike Presto watch.  It is easy to use and extremely easy to put on and take off.  T...
ASIN: B0001EKK1E, Review: I have problems with metal allergies, and I can't wear most watches. The Nike Presto watches I can w...
ASIN: B0007R4SYK, Review: I've boughten and gone through two pair of these in a row. I haven't found any that I like the style...
ASIN: B000E8F2PG, Review: I'm so disappointed with Nike.  I got this beautiful watch for my birthday and can't wear it anymore...
ASIN: B000E8H01K, Review: this watch is a durable, fantastic watch with a good reputation.  thank you amazon!!!!!!!!!!1...

=== Topic 42 ===
ASIN: B0006MFAWA, Review: i think is a really good product for a low price...
ASIN: B0007R4SYK, Review: These shoes look really nice and are very comfortable to wear. I was in need of new shoes when i bou...
ASIN: B0007RADZ8, Review: awesomeeee shoes they are really comfortable and beautiful, They are one of my best collection of sh...
ASIN: B0007RADZ8, Review: My dad loved his father day gift. Perfect size, color and style. He wears them everyday. He said it ...
ASIN: B0007RADZ8, Review: I have a lot of pairs of running shoes, and these are a nice addition to my collection. To me, they ...

=== Topic 43 ===
ASIN: B00012O2W4, Review: This watch actually looks silver when viewed online, but when I received it I saw that it was in fac...
ASIN: B0002IC3VY, Review: This is a goodlooking watch, but it doesn't have backlight as previous information detailed. I have ...
ASIN: B0006NGUE6, Review: Shoes do not fit like traditional Nike golf shoes... almost felt defective.  Ended up giving them aw...
ASIN: B0006NGUE6, Review: I had to return the shoes after wearing them just two times, because they did not live up to the wat...
ASIN: B0007RADZ8, Review: They fit right my father love them! The are very nice! They are very comfortables for walk! He like ...

=== Topic 44 ===
ASIN: B0007RADZ8, Review: green rubber is already falling off the shoe after 3 months of wearing them comfortable but glue the...
ASIN: B0007RADZ8, Review: love them they are comfortable, light weight, and the colors are great, i recommend these sneakers t...
ASIN: B000E8J52C, Review: I was looking for this color watch for months.Tried other web sites, but this watch was way over pri...
ASIN: B000J2F8GK, Review: Order these shoes for my son on January 28th and received today January 31st.coming from Taiwan. My ...
ASIN: B000J2F8GK, Review: Awesome shoes! Everything as expected specially for paying extra $50 for a pair of KD's. Will defini...

=== Topic 45 ===
ASIN: B0002QPMCI, Review: My wife wears this watch all the time and it never catches on her clothing or anything else she is r...
ASIN: B0007RADZ8, Review: always love the look of nikes but never impressed with performance of shoe.  i think they get cheape...
ASIN: B000A8BUBU, Review: This is an excelent watch. I'm really impressed with all the resources included in it. Also, the des...
ASIN: B000E8CR68, Review: I am very pleased with this product. i have been wanting it since it first came out of the market. I...
ASIN: B000G42Z2Q, Review: Item is very comfortable to use for several hours everyday, fits right to size and looks very good n...

=== Topic 46 ===
ASIN: B0006NGUE6, Review: They fit, they are fairly priced, the quality is great, they work like a charm, they have interchang...
ASIN: B0006NGUE6, Review: The shoes arrived just in time for a scheduled round of golf. The shoes were very comfortable and di...
ASIN: B0006NGUE6, Review: They were a bit stiff at first but once they got broke in they are very comfortable.  Recommend brea...
ASIN: B0007RADZ8, Review: I was a little hesitant about buying sneakers online, fearing I might end up with some fakes. But na...
ASIN: B0007RADZ8, Review: super comfy! best shoes I've bought in quite some time! nice and light, good breathability and comfo...

=== Topic 47 ===
ASIN: B0006NGUE6, Review: I have a wide foot so buying shoes over the internet is always a risk for me. These fit very nicely ...
ASIN: B0006NGUE6, Review: I had a pair just like this for like 10 years, but got sick of gluing the bottom back on. Granted I ...
ASIN: B0007RADZ8, Review: Too tight which makr it uncomfortable, I think the size may not be accurate. Esthetically it is not ...
ASIN: B0007RADZ8, Review: I bought these at Academy sports for $50. The first day they were tight and I thought I was going to...
ASIN: B0007RADZ8, Review: Shoes are too small. Understand that there is a free return.  How do we go about returning them and ...

=== Topic 48 ===
ASIN: B0002BLHCW, Review: Perfect size gym back for a typical gym locker.  Can hold a pair of sneakers, pants/shorts, towel, a...
ASIN: B0007QZ0AW, Review: I bought 2.  But  two pair both Left and right looks some different.So sad. T.Tbut very comfort. and...
ASIN: B000A8BUEC, Review: Everything was OK. I actually recommend....
ASIN: B000AI2QHW, Review: I love this bag. It's a perfect size and I use it all the time. It's also a good looking bag unlike ...
ASIN: B000AI2VE0, Review: Great bag, but i wish it were a little bigger to accomidate my gym clothes/shoes....

=== Topic 49 ===
ASIN: B0006NGUE6, Review: These shoes were very stiff when I first got them but they broke in quickly.Good pair to play in....
ASIN: B0007RADZ8, Review: bought these for supportive shoes after our daughter had ankle surgery. Great fit, great price, seem...
ASIN: B0007RADZ8, Review: Is is a good shoe, I was playing football and I needed something comfortable for when we ran sprints...
ASIN: B0007RADZ8, Review: excellent shoes, great quality and service, are as expected and i love this make are comfortable and...
ASIN: B000BY55C8, Review: I bought this as a gift for my wife - she LOVES it!!!! Everyone in my household is happy with this w...

=== Topic 50 ===
ASIN: B0006821SK, Review: First of all, battery died prematurely twice.  Both times it was replaced by a reputable jewler.  Th...
ASIN: B0006NGUE6, Review: These shoes lived up to my hopes after seeing the pictures on Amazon. The leather could be a little ...
ASIN: B0007R4U5W, Review: Very pretty smart and tough, I hope to continue shopping at amazon, I love the Nike brand, are very ...
ASIN: B0007RADZ8, Review: Stylish, great fit. Wish the one in the orange trim wasn't so expensive (and what's this all about t...
ASIN: B0007RADZ8, Review: Stylish, great fit. Wish the one in the orange trim wasn't so expensive (and what's this all about t...

=== Topic 51 ===
ASIN: B0000V9KRI, Review: good price, very good material and excellent design, very useful for traveling, totally recomendatio...
ASIN: B0006MFAWK, Review: I bought this for my 16 year old cousin, and I am sure it is going to be the perfect gift for her .....
ASIN: B0006NGUE6, Review: I purchased these for my husband, a new golfer. He says they are comfortable, and they look really s...
ASIN: B0007RADZ8, Review: very good tennis, I love the nike brand is very good quality and presentation. I recommend it. would...
ASIN: B0007RADZ8, Review: Did not fit so they have  been returned as of Nov.27 2013. Waiting for confirmation that they have r...

=== Topic 52 ===
ASIN: B000BJVY5K, Review: If you're an air max fan, I recommend these ones. I love my old air max's and it was time for a new ...
ASIN: B000E8H01K, Review: The "Nike Men's Hammer Watch" (#C0021-001) is freakin' AWESOME!!! It is pretty huge but I really lik...
ASIN: B000E8H0BU, Review: I have had difficulty finding a watch for swimming.  This has met all my needs.  Of good size so eas...
ASIN: B000G42Z2Q, Review: I've owned many Nikes before, but have never had a sneaker hurt the top of my foot like this. In the...
ASIN: B000OFN364, Review: The Nike TRIAX Air Structure 10 I purchased had a very poorly made left shoe, leaning far to the lef...

=== Topic 53 ===
ASIN: B0001YMVSY, Review: It is oversized for my wrist. It will be fine to wear around but not for running which is what I bou...
ASIN: B0007RADZ8, Review: Fit as expected. So far they are pretty comfortable to run in which is key for me getting out there....
ASIN: B0007RADZ8, Review: I can never find cool looking sneakers in wides in 13's.Oh wait i did.These are nice just fit snug....
ASIN: B000A8BTMK, Review: I bought this as a gift for my youngest daughter's birthday. She absolutely loves it! Of course my o...
ASIN: B000A8BTQ6, Review: Tells time with giant numbers on screen. Rocks my socks off....

=== Topic 54 ===
ASIN: B0007R6CAS, Review: My lady got these as a gift.  I really like these shoes. I will drop hints about another pair I like...
ASIN: B0007RADZ8, Review: It was great and my son was happy with his new Nike revolution shoes 10.5 men white color bought in ...
ASIN: B0007RADZ8, Review: The entire transaction was great, easy to order, quick to ship and arrived faster than expected.  I'...
ASIN: B000E8H01K, Review: This was a gift, which the person loved, and I loved the price, $30.00 cheaper than everywhere else...
ASIN: B000P6GKOO, Review: Great spring/summer shoe. Not really for running. Fits true to size but is also a soft mesh so there...

=== Topic 55 ===
ASIN: B00012O2O2, Review: This watch is ideal for a person with a very small wrist.  Looks very nice but is light green not ca...
ASIN: B00012O2RY, Review: Hello good evening, excellent product quality, comfortable, nice, good wreck 100% recommend them ver...
ASIN: B00019WP6M, Review: This shirt is of good quality and logo of the team is large and striking....
ASIN: B0007RADZ8, Review: excellent quality, are accurate to the picture showing, Super Light, wonderful finish, ideal for wal...
ASIN: B000E8CRZ4, Review: this was a great watch, excellent service.  The order arrived on time and I love the watch....

=== Topic 56 ===
ASIN: B00012O2LA, Review: I own three of these watches. They are perfect for an active lifestyle, water resistant, a chron to ...
ASIN: B00012O2O2, Review: I had great hopes for this watch as I wanted a funky cool change for spring-- the colors do not cont...
ASIN: B00012O2R4, Review: Very nice watch for a teenager, it has a pretty blue backlight and looks so nice when you wear it....
ASIN: B00019FWE4, Review: I'm on my 4th watch... I keep returning it due to poor design.  The band keeps coming apart in the s...
ASIN: B0001YMVTS, Review: Im happy with this watch so far. In my type of work I always need a durable watch. I used to be a bi...

=== Topic 57 ===
ASIN: B0007RADZ8, Review: fit somewhat tight but as expected! great buy, fast shipping and cheap price....
ASIN: B0007RADZ8, Review: The price was perfect and the shoes are very confortable. The packaging was in perfect conditions. I...
ASIN: B0007RADZ8, Review: I have found men's sneakers fit me better than women's (I wear a 13.5 or 14 in women's - I'm 6 feet ...
ASIN: B0007RADZ8, Review: No where as good as the Nike air I bought locally for less. Very disappointed  in the shoe.  I have ...
ASIN: B000A8BU4M, Review: This watch is perfect for exercise, small, reliable, and has a lot option that you select. I recomme...

=== Topic 58 ===
ASIN: B0007RADZ8, Review: Since I purchased goods I need in Amazon, I think everything is good except COMBO SET YONEX Badminto...
ASIN: B0007RADZ8, Review: I got these for my husband and he liked them. Theyre a lot lower than expected at the ankle, He thou...
ASIN: B0007RADZ8, Review: These are a good looking pair of shoes, they fit well and feel pretty good and for a good price. My ...
ASIN: B000E8F1WK, Review: I like this watch however it is a bit to thick for my wrist. When I wear it, the band gets in the wa...
ASIN: B000FAG88S, Review: This bag is actually larger than I thought.  You can really fit a lot into it.  This would be good f...

=== Topic 59 ===
ASIN: B0000V9K3W, Review: The watch was exactly what i ordered and I got it very fast. Unfortunately it was a bit too big for ...
ASIN: B0006NGUE6, Review: Great looking and really good feeling shoes.  They are stable and comfortable.  It's like playing in...
ASIN: B0006NGUE6, Review: Sent wrong color shoes. Sent white when I ordered black but I had to use them anyway because of time...
ASIN: B0007RADZ8, Review: The size was not right, but I would have returned them anyway as they arrived laced up, with sock fu...
ASIN: B0007RADZ8, Review: Great shoes for the price, I work out 5 days a week doing weights and cardio. They are comfortable a...

=== Topic 60 ===
ASIN: B0001YMVSY, Review: Great watch, very comfortable, good size , it fits nice, good color, if you have bigger wrist be awa...
ASIN: B0002164KC, Review: By far, the best pair of shoes I've ever owned. Well, you only need to know that their size is norma...
ASIN: B0006MFAW0, Review: This watch is very cute. It is described as being a petite watch but that is a polite word for sayin...
ASIN: B0006NGUE6, Review: I received these shoes last week. Aside from the usual break in period (stiffness), these shoes didn...
ASIN: B0006NGUE6, Review: a good looking shoe but not to be recommended to buy unseen and untried. For the size on the shoe, i...

=== Topic 61 ===
ASIN: B0006NGUE6, Review: As comfortable if nor more than foot joy dryjoys. Have played a few rounds no issues.  Might have to...
ASIN: B000AYJLUQ, Review: This is the most comfortable mat that I have ever used. Has a great surface and I love the dual colo...
ASIN: B000G42Z2Q, Review: I use these Nike Shoes for working out at the Gym. they feel lite and comfortable. They fit very wel...
ASIN: B000J2F8GK, Review: This is a great shoe to either play basketball in or to wear with your street clothes I personally l...
ASIN: B000P6FKAE, Review: As a sneaker collector, I am always on the prowl for Nike Dunks, so I was extremely happy to find th...

=== Topic 62 ===
ASIN: B0007RADZ8, Review: I am pretty active after my 9-5 with wight lifting, running, circuit traing and these shoes have bee...
ASIN: B0007RADZ8, Review: They are great shoes for me to where anywhere and work in them numerous days a week.  They are nice ...
ASIN: B000AORSVU, Review: I'm very happy with my purchase. I where my sunnies everyday. I really like this style. I will buy t...
ASIN: B000P6GKOO, Review: after a couple of months wearing them the suede pulled out of the sole and there is a blow-out hole ...
ASIN: B000V4YZ1K, Review: I ordered this item for a co- worker son. She wanted to get them something but don't have a computer...

=== Topic 63 ===
ASIN: B0001EKJZG, Review: i had a brown one and wanted another. like the color purple a lot but would prefer the wristband "ho...
ASIN: B0007RADZ8, Review: Not too big and not to small. Looks like not as the same as I expected, but I think it is not big de...
ASIN: B000G42Z2Q, Review: These are, naturally, a good quality shoe, coming from Nike. They fit true to size and are very comf...
ASIN: B000P6GKOO, Review: Great shoe for general use, not a real running shoe....very comfortable with or w/o socks. Perfect s...
ASIN: B000P6GKOO, Review: PERFECT EVERYTHING!!!! THESE SHOES IS MY MUZZAFELKIN NEW FREEINDS YO. IF UZ UP 2 ME ID OWNZ LYK 200 ...

=== Topic 64 ===
ASIN: B0000V9K32, Review: the colour i received is not blue as shown but yellow.Couldnt change it because it was a birthday pr...
ASIN: B0001EKJZG, Review: I love this watch and have it in a couple of other colors - when I saw I could get a pink one, I fig...
ASIN: B0007RADZ8, Review: They are very comfortable and I've gotten lots of compliments from people. If you're looking for a g...
ASIN: B0007RADZ8, Review: definitively I have not received what I needed color, color were wrong order. when he received my su...
ASIN: B0007RADZ8, Review: It's like walking on air. These are so comfortable to wear and I love running in these! I've gotten ...

=== Topic 65 ===
ASIN: B00012O2O2, Review: I've had several of these watches in various colors and love them. The picture is very misleading.  ...
ASIN: B00012O2RE, Review: I purchased this for one of my teenage daughters as she has small wrists and really liked the look o...
ASIN: B00012O2RY, Review: This is a beautiful watch, bigger than I thought it would be, which made me happy. The back light wo...
ASIN: B00012O2SI, Review: I love this watch.  Easy to wear to work and for every day. It is a great watch.  I wish Nike still ...
ASIN: B00012O2SI, Review: I am allergic to metal clasps so this bracelet type works well for me.  This Nike line was discontin...

=== Topic 66 ===
ASIN: B0006NGUE6, Review: Very nice and comfortable golf shoes!! This was the only place I could find these to fit me. Will lo...
ASIN: B0006NGUE6, Review: I'm a total amateur that's starting to understand the way the game is played, instead of just swingi...
ASIN: B0007RADZ8, Review: Nice shoes!...
ASIN: B0007RADZ8, Review: Very nice shoe and comfortable....
ASIN: B000E8J5N6, Review: This product is great for helping beginners improve their running, but it is not for the techno-phob...

=== Topic 67 ===
ASIN: B0001YMVVG, Review: The watch is great! Came in a few days and was in brand new condition, as expected....
ASIN: B0006NGUE6, Review: Very pleased with how easy it was.  Purchase arrived in a very timely fashion...shoes fit very well ...
ASIN: B0007RADZ8, Review: Just as expected. This pair is perfect for work and are very comfortable. Great price and great prod...
ASIN: B0007RADZ8, Review: My husband said they are very comfortable and easy to walk in, which he does a lot of. They also loo...
ASIN: B0007RADZ8, Review: They are very comfortable, they look exactly the same in the picture. My son was very happy and will...

=== Topic 68 ===
ASIN: B0000V9KRI, Review: I love this watch, i use every day, every where, to work, to go to the gym, pool, sports, etc. Eveth...
ASIN: B0001EKJZG, Review: My GF loves purple and works in the NICU. This watch is perfect for that and she loves it. It comes ...
ASIN: B0002IC3ZU, Review: I like the angle; makes it much easier to read.  Numbers don't need to be stretched - get the one w/...
ASIN: B00073G91Y, Review: Comfortable fit.  Wore in the ocean and pool and worked great...
ASIN: B0007RADZ8, Review: Got em for my hubbys birthday. He loves the darn things, super light weight, comfortable, easy to wa...

=== Topic 69 ===
ASIN: B0007TQLC0, Review: I love this watch and I wear it from Memorial Day to Labor Day.  I wish I could find it in clear....
ASIN: B000A8BTPC, Review: I can not give a positive or negative opinion about the watch, as up to today I have NOT received it...
ASIN: B000A8BU0Q, Review: This watch has been literally all over the world from the top of mountains to the bottom of the ocea...
ASIN: B000E8J5B8, Review: This is quite possibly the best watch I've ever worn. It looks really stylish and everyone that sees...
ASIN: B000G42Z2Q, Review: Amazing shoes, good for running, going to the gym and any physical exercise.The price was excellent ...

=== Topic 70 ===
ASIN: B00019WPTE, Review: This was a bday gift for one of my family members, who is of course a flyers fan. It got to him in g...
ASIN: B0001YMVVG, Review: This watch is sporty, lightweight, easy to take on and off.  I'm a nurse and wash my hands often so ...
ASIN: B0006NGUE6, Review: These fit perfect and no break in period. I'm very happy with the comfort and looks of this shoe. Th...
ASIN: B0007RADZ8, Review: After wearing these for about a month I can say a few things:Pros:- The shoes look fantastic and app...
ASIN: B0007RADZ8, Review: great fit, comfortable, delivered on time. was looking for a comfortable gym shoe for my 15 year old...

=== Topic 71 ===
ASIN: B0006NGUE6, Review: They are confortable to wear and are of quality construction. They haven't helped my golf game thoug...
ASIN: B0006NGUE6, Review: shoes fit great. exact color I wanted and it helped my golf game a lot and helped me drive 320 yards...
ASIN: B0009RSNCQ, Review: Perfect fit, look, quality. Great for every day use including jogging or sports.So comfortable I for...
ASIN: B000G42Z2Q, Review: When I recieved this item one of the silver lace loops (Which you can see in the picture) was ripped...
ASIN: B000J2F8GK, Review: My son loves the shoes! Though we ordered them a bit too big....he still loves them and gets plenty ...

=== Topic 72 ===
ASIN: B000A8BU38, Review: This watch is very light and moves well with you in high action sports.I like the design of the band...
ASIN: B000AI5NRW, Review: great bag. i have had this for a couple of years now and it has held up wonderfully. wouldn't expect...
ASIN: B000P6GKOO, Review: -----The Good Stuff------great ventilation and cushioning-lightweight impact protection-really versa...
ASIN: B000V4YZ1K, Review: the sandals fit great. But, wrong size. Ineed a 12 size to fit me.  the 16 is longer than i need the...
ASIN: B0013UXIAK, Review: You need to ask my child about these shoes..... wears them all the time, they do not leave his room ...

=== Topic 73 ===
ASIN: B0006NGUE6, Review: I purchased these as a gift for my son.  He loved them.  They fit well, he thought they were well ma...
ASIN: B0007RADZ8, Review: Not to tight fitting, still had space to move foot around. Lightweight with enough air circulation t...
ASIN: B000G3PT1G, Review: I play freshman football and I needed something to add to my uniform. These were perfect. The color ...
ASIN: B000UQL1T8, Review: These boots are terrible! Yes they are stylish and comfortable but they are not waterproof what so e...
ASIN: B000V4YZ1K, Review: Tight across the arches. Will hopefully stretch with time. Otherwise, they have very good constructi...

=== Topic 74 ===
ASIN: B0000V9KNM, Review: Why isn't Nike making these anymore?  I love this watch, and I get a lot of compliments, questions f...
ASIN: B0001YMVVQ, Review: This watch is great because it's different, comfortable, and simple.  It doesn't have a timer or any...
ASIN: B000E8H060, Review: Swanky sporting watch..got a true fit.I even purchased the gold its slightly different but just as n...
ASIN: B000E8H0BU, Review: My dad wanted this watch for working out and playing sports after he saw it at Dicks Sporting Goods....
ASIN: B000G3PT1G, Review: not so good as i expected, it doesn't have the lable as the picture shows and extra strings are obvi...

# ==========================
# 10. SAVE RESULTS
# ==========================
reviews_df.to_csv("nike_reviews_with_topics.csv", index=False)

df1= pd.read_csv("nike_reviews_with_topics.csv")
df1

# ==========================
# 9. INSIGHTS
# ==========================
# Example: Most common ASINs per topic
topic_summary = reviews_df.groupby('topic')['asin'].value_counts().groupby(level=0).head(3)

# Convert to DataFrame and give count column a proper name
result = topic_summary.sort_values(ascending=False).head(25).reset_index(name='count')

# Set pandas display options to show full text
pd.set_option('display.max_colwidth', None)
pd.set_option('display.width', None)
pd.set_option('display.max_columns', None)

# Display the result
result

# ==========================
# 9. INSIGHTS
# ==========================
# Top ASIN (most reviewed product) for each topic
topic_summary = reviews_df.groupby('topic')['asin'].value_counts().groupby(level=0).head(1)

# Convert to DataFrame and give count column a proper name
result = topic_summary.reset_index(name='review_count')

# Set pandas display options to show full text
pd.set_option('display.max_colwidth', None)
pd.set_option('display.width', None)
pd.set_option('display.max_columns', None)

# Display the result
result

# ==========================
# INTERACTIVE PLOT WITH TOOLTIPS
# ==========================
import plotly.express as px
import plotly.graph_objects as go
import pandas as pd

# Get the top ASIN for each topic
topic_summary = reviews_df.groupby('topic')['asin'].value_counts().groupby(level=0).head(1)
result = topic_summary.reset_index(name='review_count')

# Get sample review text for each top ASIN to show in tooltip
def get_sample_reviews(topic, asin, n_samples=3):
    sample_reviews = reviews_df[
        (reviews_df['topic'] == topic) & 
        (reviews_df['asin'] == asin)
    ]['reviewText'].head(n_samples).tolist()
    
    # Truncate long reviews and join with line breaks
    truncated_reviews = []
    for review in sample_reviews:
        if len(str(review)) > 150:
            truncated_reviews.append(str(review)[:150] + "...")
        else:
            truncated_reviews.append(str(review))
    
    return "<br><br>".join(truncated_reviews)

# Add sample review texts to the result dataframe
result['sample_reviews'] = result.apply(
    lambda row: get_sample_reviews(row['topic'], row['asin']), axis=1
)

# Create interactive bar plot
fig = px.bar(
    result, 
    x='topic', 
    y='review_count',
    title='Top Product (ASIN) by Review Count per Topic',
    labels={'review_count': 'Number of Reviews', 'topic': 'Topic'},
    color='review_count',
    color_continuous_scale='viridis'
)

# Update traces to include custom hover template
fig.update_traces(
    hovertemplate="<b>Topic:</b> %{x}<br>" +
                  "<b>ASIN:</b> %{customdata[0]}<br>" +
                  "<b>Review Count:</b> %{y}<br><br>" +
                  "<b>Sample Reviews:</b><br>%{customdata[1]}" +
                  "<extra></extra>",
    customdata=list(zip(result['asin'], result['sample_reviews']))
)

# Customize layout
fig.update_layout(
    xaxis_tickangle=-45,
    height=600,
    width=1000,
    showlegend=False,
    hovermode='closest'
)

# Show the plot
fig.show()

# ==========================
# INTERACTIVE PLOT WITH TOP 2 PRODUCTS AND TOOLTIPS
# ==========================
import plotly.express as px
import pandas as pd

# Get the top 2 ASINs for each topic
topic_summary = (
    reviews_df.groupby('topic')['asin']
    .value_counts()
    .groupby(level=0)
    .head(3)  # top 2 products per topic
    .reset_index(name='review_count')
)

# Function to get sample reviews for a given topic and ASIN
def get_sample_reviews(topic, asin, n_samples=3):
    sample_reviews = reviews_df[
        (reviews_df['topic'] == topic) & 
        (reviews_df['asin'] == asin)
    ]['reviewText'].head(n_samples).tolist()
    
    truncated_reviews = []
    for review in sample_reviews:
        if len(str(review)) > 150:
            truncated_reviews.append(str(review)[:150] + "...")
        else:
            truncated_reviews.append(str(review))
    return "<br><br>".join(truncated_reviews)

# Add sample review texts to the dataframe
topic_summary['sample_reviews'] = topic_summary.apply(
    lambda row: get_sample_reviews(row['topic'], row['asin']), axis=1
)

# Create a combined tooltip string for each topic showing both top products
tooltip_data = []
for topic in topic_summary['topic'].unique():
    subset = topic_summary[topic_summary['topic'] == topic]
    hover_text = ""
    for idx, row in subset.iterrows():
        hover_text += f"<b>Topic:</b> {row['topic']}<br>"
        hover_text += f"<b>ASIN:</b> {row['asin']}<br>"
        hover_text += f"<b>Review Count:</b> {row['review_count']}<br>"
        hover_text += f"<b>Sample Reviews:</b><br>{row['sample_reviews']}<br><br>"
    tooltip_data.append((topic, hover_text.strip()))

# Prepare dataframe for plotting: one bar per topic (sum of top 2 review counts)
plot_df = topic_summary.groupby('topic')['review_count'].sum().reset_index()
plot_df['hover_text'] = plot_df['topic'].map(dict(tooltip_data))

# Create interactive bar plot
fig = px.bar(
    plot_df,
    x='topic',
    y='review_count',
    title='Top 3 Products by Review Count per Topic',
    labels={'review_count': 'Sum of Top 3 Reviews', 'topic': 'Topic'},
    color='review_count',
    color_continuous_scale='viridis'
)

# Update hover template
fig.update_traces(
    hovertemplate="%{customdata}<extra></extra>",
    customdata=plot_df['hover_text']
)

# Customize layout
fig.update_layout(
    xaxis_tickangle=-45,
    height=600,
    width=1000,
    showlegend=False,
    hovermode='closest'
)

fig.show()

import pandas as pd
import plotly.express as px

# Example: Aggregate topic counts and average rating
topic_summary = (
    reviews_df.groupby('topic')
    .agg(
        count=('topic', 'size'),
        avg_rating=('overall', 'mean'),
        topic_top_words=('topic_top_words', 'first')
    )
    .reset_index()
)

# Plot distribution with tooltips
fig = px.bar(
    topic_summary,
    x='topic',
    y='count',
    text='count',
    hover_data={'avg_rating': ':.2f', 'topic': True, 'count': True, 'topic_top_words':True},
    labels={'topic': 'Topic', 'count': 'Number of Reviews'},
    title='Distribution of 75 Topics in Reviews with Average Rating'
)

fig.update_traces(marker_color='steelblue')
fig.update_layout(xaxis=dict(tickmode='linear'))
fig.show()

import plotly.graph_objects as go
from plotly.subplots import make_subplots

# Sort top and bottom 5 topics
top5 = topic_summary.sort_values(by="count", ascending=False).head(5)
bot5 = topic_summary.sort_values(by="count", ascending=True).head(5)

# Create subplot figure with 1 row, 2 columns
fig = make_subplots(rows=1, cols=2, specs=[[{'type':'domain'}, {'type':'domain'}]],
                    subplot_titles=('Top 5 Topics', 'Bottom 5 Topics'))

# Add top 5 pie
fig.add_trace(
    go.Pie(
        labels=top5['topic'],
        values=top5['count'],
        hovertext=top5['topic_top_words'],
        hoverinfo='label+value+text'
    ),
    row=1, col=1
)

# Add bottom 5 pie
fig.add_trace(
    go.Pie(
        labels=bot5['topic'],
        values=bot5['count'],
        hovertext=bot5['topic_top_words'],
        hoverinfo='label+value+text'
    ),
    row=1, col=2
)

# Update layout
fig.update_layout(title_text='Top and Bottom 5 Topics by Count in Reviews')
fig.show()

import pandas as pd
from nltk.sentiment import SentimentIntensityAnalyzer
import matplotlib.pyplot as plt
import seaborn as sns

# Make sure NLTK vader lexicon is downloaded
import nltk
nltk.download('vader_lexicon')

sia = SentimentIntensityAnalyzer()

# Compute sentiment score for each review
reviews_df['sentiment'] = reviews_df['reviewText'].apply(lambda x: sia.polarity_scores(x)['compound'])

[nltk_data] Downloading package vader_lexicon to
[nltk_data]     C:\Users\forca\AppData\Roaming\nltk_data...
[nltk_data]   Package vader_lexicon is already up-to-date!

# Aggregate both average sentiment and overall rating per topic
topic_summary = reviews_df.groupby('topic').agg(
    avg_sentiment=('sentiment', 'mean'),
    avg_overall=('overall', 'mean'),
    count=('reviewText', 'count')
).reset_index()

# Interactive bar plot
fig = px.bar(
    topic_summary,
    x='topic',
    y='avg_sentiment',
    color='avg_sentiment',  # diverging color map
    color_continuous_scale='RdBu_r',  # red=positive, blue=negative
    hover_data={
        'avg_sentiment': ':.2f',
        'avg_overall': ':.2f',
        'count': True
    },
    labels={'avg_sentiment': 'Avg Sentiment', 'topic': 'Topic'},
    title='Average Sentiment per Topic with Avg Overall Rating'
)

fig.update_layout(xaxis_tickangle=-45)
fig.show()

# ==========================
# ADVANCED ANALYSIS IDEAS
# ==========================

# 1. SEGMENTATION BY STAR RATINGS (Easy)
# =====================================
print("1. Analysis by Star Ratings")
print("="*40)

# Separate positive vs negative reviews
positive_reviews = reviews_df[reviews_df['overall'] >= 4]
negative_reviews = reviews_df[reviews_df['overall'] <= 2]
neutral_reviews = reviews_df[reviews_df['overall'] == 3]

# Compare topic distributions
positive_topics = positive_reviews['topic'].value_counts().head(5)
negative_topics = negative_reviews['topic'].value_counts().head(5)

print("Top topics in positive reviews:")
print(positive_topics)
print("\nTop topics in negative reviews:")  
print(negative_topics)

# Visualize the comparison
import matplotlib.pyplot as plt
import seaborn as sns

fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(15, 6))
positive_topics.plot(kind='barh', ax=ax1, title='Top Topics - Positive Reviews (4-5 stars)')
negative_topics.plot(kind='barh', ax=ax2, title='Top Topics - Negative Reviews (1-2 stars)')
plt.tight_layout()
plt.show()

1. Analysis by Star Ratings
========================================
Top topics in positive reviews:
topic
60    1074
67     781
27     616
42     500
37     469
Name: count, dtype: int64

Top topics in negative reviews:
topic
60    243
37    173
27    141
33     79
47     79
Name: count, dtype: int64

# ==========================================
# WORD FREQUENCIES PER TOPIC (POS vs NEG)
# ==========================================

from collections import Counter
import matplotlib.pyplot as plt

def get_top_words(df, topic_id, sentiment_label, n=20):
    """
    df: reviews dataframe
    topic_id: int, topic number
    sentiment_label: 'positive' or 'negative'
    n: top N words
    """
    if sentiment_label == "negative":
        subset = df[(df['topic'] == topic_id) & (df['overall'] <= 2)]
    elif sentiment_label == "positive":
        subset = df[(df['topic'] == topic_id) & (df['overall'] >= 4)]
    else:
        raise ValueError("sentiment_label must be 'positive' or 'negative'")
    
    # Flatten all tokens from 'cleaned' (assumes it's tokenized list or space-separated string)
    words = []
    for text in subset['cleaned']:
        if isinstance(text, str):
            words.extend(text.split())
        elif isinstance(text, list):
            words.extend(text)
    
    return Counter(words).most_common(n)


# Define target topics
negative_topics_list = [53, 63, 51, 19, 14]
positive_topics_list = [63, 14, 70, 19, 64]

# Collect results
topic_word_stats = {}

for t in negative_topics_list:
    topic_word_stats[(t, 'negative')] = get_top_words(reviews_df, t, 'negative', n=20)

for t in positive_topics_list:
    topic_word_stats[(t, 'positive')] = get_top_words(reviews_df, t, 'positive', n=20)


# ========== DISPLAY RESULTS (commented out the print) ==========
#for (topic_id, sentiment), words in topic_word_stats.items():
#    print(f"\nTop words for topic {topic_id} ({sentiment} reviews):")
#    for word, count in words:
#        print(f"{word}: {count}")

fig, axes = plt.subplots(2, 5, figsize=(25, 10))
axes = axes.flatten()

for idx, ((topic_id, sentiment), words) in enumerate(topic_word_stats.items()):
    labels, values = zip(*words) if words else ([], [])
    axes[idx].barh(labels, values)
    axes[idx].set_title(f"Topic {topic_id} ({sentiment})")
    axes[idx].invert_yaxis()

plt.tight_layout()
plt.show()

# Define target topics
negative_topics_list = [63, 64, 14, 0, 25]
positive_topics_list = [63, 64, 14, 0, 25]

# Collect results
topic_word_stats = {}

for t in negative_topics_list:
    topic_word_stats[(t, 'negative')] = get_top_words(reviews_df, t, 'negative', n=20)

for t in positive_topics_list:
    topic_word_stats[(t, 'positive')] = get_top_words(reviews_df, t, 'positive', n=20)


# ========== DISPLAY RESULTS (commented out the print) ==========
#for (topic_id, sentiment), words in topic_word_stats.items():
#    print(f"\nTop words for topic {topic_id} ({sentiment} reviews):")
#    for word, count in words:
#        print(f"{word}: {count}")

fig, axes = plt.subplots(2, 5, figsize=(25, 10))
axes = axes.flatten()

for idx, ((topic_id, sentiment), words) in enumerate(topic_word_stats.items()):
    labels, values = zip(*words) if words else ([], [])
    axes[idx].barh(labels, values)
    axes[idx].set_title(f"Topic {topic_id} ({sentiment})")
    axes[idx].invert_yaxis()

plt.tight_layout()
plt.show()

import plotly.graph_objs as go
from plotly.subplots import make_subplots

# topic_word_stats: dict keyed by (topic_id, sentiment) -> [(word, count), ...]
pairs = list(topic_word_stats.items())  # preserves ordering
rows, cols = 5, 2                       # 5 rows x 2 cols = 10 slots

# Titles for each subplot
titles = [f"Topic {tid} ({sent})" for (tid, sent), _ in pairs]

fig = make_subplots(
    rows=rows, cols=cols,
    subplot_titles=titles,
    horizontal_spacing=0.15,
    vertical_spacing=0.1
)

# Add the 10 plots into the grid (row-major order)
for idx, ((topic_id, sentiment), words) in enumerate(pairs):
    row = idx // cols + 1
    col = idx % cols + 1

    if words:
        labels, values = zip(*words)
        labels, values = list(labels), list(values)
    else:
        labels, values = [], []

    fig.add_trace(
        go.Bar(
            x=values,
            y=labels,
            orientation='h',
            hovertemplate="Word: %{y}<br>Count: %{x}<extra></extra>"
        ),
        row=row, col=col
    )

    # Invert y-axis (like matplotlib barh)
    fig.update_yaxes(
        autorange="reversed",
        automargin=True,
        tickfont=dict(size=10),
        row=row, col=col
    )
    fig.update_xaxes(automargin=True, row=row, col=col)

# Layout tweaks
fig.update_layout(
    height=1800,
    width=1200,
    title_text="Top Words by Topic (Positive vs Negative) — 2 cols × 5 rows",
    showlegend=False,
    margin=dict(l=100, r=60, t=100, b=60)
)

fig.show()

import plotly.graph_objs as go
from plotly.subplots import make_subplots
import random

print("positive is 4+ stars, negative is 2- stars.")
# -------------------------------
# Define topic groups
# -------------------------------
negative_topic_ids = [53, 63, 51, 19, 14]
positive_topic_ids = [63, 14, 70, 19, 64]

# -------------------------------
# Build filtered ASIN counts and truncated sample reviews
# -------------------------------
all_topic_asins = {}

def get_top_asins_with_reviews(subset, n_asins=5, n_reviews=3, max_len=150):
    top_asins = subset['asin'].value_counts().head(n_asins)
    hover_texts = []

    for asin in top_asins.index:
        reviews = subset[subset['asin'] == asin]['reviewText'].dropna().tolist()
        sample_reviews = random.sample(reviews, min(len(reviews), n_reviews))
        
        # Truncate long reviews
        truncated_reviews = []
        for review in sample_reviews:
            review_str = str(review)
            if len(review_str) > max_len:
                truncated_reviews.append(review_str[:max_len] + "...")
            else:
                truncated_reviews.append(review_str)
        
        hover_text = "<br>".join(truncated_reviews)
        hover_texts.append(hover_text)
    
    return top_asins, hover_texts

# Negative topics
for t in negative_topic_ids:
    subset = reviews_df[(reviews_df['topic'] == t) & (reviews_df['overall'] <= 2)]
    top_asins, hover_texts = get_top_asins_with_reviews(subset)
    all_topic_asins[(t, 'Negative')] = (top_asins, hover_texts)

# Positive topics
for t in positive_topic_ids:
    subset = reviews_df[(reviews_df['topic'] == t) & (reviews_df['overall'] >= 4)]
    top_asins, hover_texts = get_top_asins_with_reviews(subset)
    all_topic_asins[(t, 'Positive')] = (top_asins, hover_texts)

# -------------------------------
# Prepare subplot layout
# -------------------------------
pairs = list(all_topic_asins.items())
rows, cols = 5, 2  # 10 topics -> 2 cols x 5 rows
titles = [f"Topic {tid} ({sentiment})" for (tid, sentiment), _ in pairs]

fig = make_subplots(
    rows=rows, cols=cols,
    subplot_titles=titles,
    horizontal_spacing=0.15,
    vertical_spacing=0.1
)

# -------------------------------
# Add ASIN bar charts with truncated sample reviews in hover
# -------------------------------
for idx, ((topic_id, sentiment), (asin_counts, hover_texts)) in enumerate(pairs):
    row = idx // cols + 1
    col = idx % cols + 1

    labels = asin_counts.index.tolist()
    values = asin_counts.values.tolist()

    fig.add_trace(
        go.Bar(
            x=values,
            y=labels,
            orientation='h',
            hovertemplate=[
                f"ASIN: {asin}<br>Count: {count}<br><br>Sample Reviews:<br>{hover}"
                for asin, count, hover in zip(labels, values, hover_texts)
            ]
        ),
        row=row, col=col
    )

    fig.update_yaxes(autorange="reversed", automargin=True, tickfont=dict(size=10), row=row, col=col)
    fig.update_xaxes(automargin=True, row=row, col=col)

# -------------------------------
# Layout tweaks
# -------------------------------
fig.update_layout(
    height=1800,
    width=1200,
    title_text="Top 5 ASINs per Topic (Filtered by Negative/Positive Reviews) with Truncated Sample Reviews",
    showlegend=False,
    margin=dict(l=100, r=60, t=100, b=60)
)

fig.show()

positive is 4+ stars, negative is 2- stars.

import random
import plotly.graph_objects as go
from plotly.subplots import make_subplots

print("positive is 4+ stars, negative is 2- stars.")

# -------------------------------
# Define topic groups
# -------------------------------
topic_ids = [63, 64, 14, 0, 25]

# -------------------------------
# Function to get top ASINs and sample reviews
# -------------------------------
def get_top_asins_with_reviews(subset, n_asins=5, n_reviews=3, max_len=150):
    top_asins = subset['asin'].value_counts().head(n_asins)
    hover_texts = []

    for asin in top_asins.index:
        reviews = subset[subset['asin'] == asin]['reviewText'].dropna().tolist()
        if not reviews:
            hover_texts.append("No reviews available")
            continue
        sample_reviews = random.sample(reviews, min(len(reviews), n_reviews))

        truncated_reviews = []
        for review in sample_reviews:
            review_str = str(review)
            if len(review_str) > max_len:
                truncated_reviews.append(review_str[:max_len] + "...")
            else:
                truncated_reviews.append(review_str)

        hover_text = "<br>".join(truncated_reviews)
        hover_texts.append(hover_text)

    return top_asins, hover_texts

# -------------------------------
# Collect data for each topic/sentiment
# -------------------------------
all_topic_asins = {}

for t in topic_ids:
    # Negative reviews
    subset_neg = reviews_df[(reviews_df['topic'] == t) & (reviews_df['overall'] <= 2)]
    top_asins_neg, hover_neg = get_top_asins_with_reviews(subset_neg)
    all_topic_asins[(t, 'Negative')] = (top_asins_neg, hover_neg)

    # Positive reviews
    subset_pos = reviews_df[(reviews_df['topic'] == t) & (reviews_df['overall'] >= 4)]
    top_asins_pos, hover_pos = get_top_asins_with_reviews(subset_pos)
    all_topic_asins[(t, 'Positive')] = (top_asins_pos, hover_pos)

# -------------------------------
# Prepare subplot layout
# -------------------------------
rows = len(topic_ids)
cols = 2  # Negative | Positive
titles = []
for t in topic_ids:
    titles.extend([f"Topic {t} (Negative)", f"Topic {t} (Positive)"])

fig = make_subplots(
    rows=rows, cols=cols,
    subplot_titles=titles,
    horizontal_spacing=0.15,
    vertical_spacing=0.15
)

# -------------------------------
# Add ASIN bar charts
# -------------------------------
for row_idx, t in enumerate(topic_ids, start=1):
    for col_idx, sentiment in enumerate(['Negative', 'Positive'], start=1):
        asin_counts, hover_texts = all_topic_asins[(t, sentiment)]
        labels = asin_counts.index.tolist()
        values = asin_counts.values.tolist()

        fig.add_trace(
            go.Bar(
                x=values,
                y=labels,
                orientation='h',
                hovertemplate=[
                    f"ASIN: {asin}<br>Count: {count}<br><br>Sample Reviews:<br>{hover}"
                    for asin, count, hover in zip(labels, values, hover_texts)
                ]
            ),
            row=row_idx, col=col_idx
        )

        fig.update_yaxes(autorange="reversed", automargin=True, tickfont=dict(size=10), row=row_idx, col=col_idx)
        fig.update_xaxes(automargin=True, row=row_idx, col=col_idx)

# -------------------------------
# Layout tweaks
# -------------------------------
fig.update_layout(
    height=1800,
    width=1200,
    title_text="Top 5 ASINs per Topic (Negative vs Positive) with Sample Reviews",
    showlegend=False,
    margin=dict(l=100, r=60, t=100, b=60)
)

fig.show()

positive is 4+ stars, negative is 2- stars.

# 3. TEMPORAL ANALYSIS 
# ==================================
print("\n3. Temporal Analysis")
print("="*40)

# Convert reviewTime to datetime if not already
reviews_df['reviewTime'] = pd.to_datetime(reviews_df['reviewTime'])
reviews_df['year'] = reviews_df['reviewTime'].dt.year
reviews_df['month'] = reviews_df['reviewTime'].dt.month

# Seasonal analysis
seasonal_topics = reviews_df.groupby(['month', 'topic']).size().unstack(fill_value=0)
print("\nTopic trends by month (seasonality):")
print(seasonal_topics.head(12))

3. Temporal Analysis
========================================

Topic trends by month (seasonality):
topic  0   1   2   3   4   5   6   7   8   9   10  11  12  13  14  15  16  17  \
month                                                                           
1       4  30  16  22  34  34  28  59  12  35  21  15  23  26  39  30  16   8   
2       5  19  23  21  15  18  25  26  14  38  18  11  24  39  40  36   9   8   
3       8  38  22  22  35  25  38  46  18  36  14  14  17  33  40  31  17  12   
4       3  24  38  22  28  36  24  53  31  16  19  22  21  47  52  40  22  13   
5       8  41  22  24  30  36  24  38  24   5  11  17  18  42  50  29  15  11   
6      12  32  32  23  27  28  29  42  33  12  15  15  13  34  43  42  18   7   
7      29  24  27  21  25  21  23  46  15  23  18   9  40  36  49  45   8  13   
8       3  12  14   9  20  16  13  15  11   5   8   7   8  18  23  11   8   3   
9       2  15  10  15  15  14  14  21   7   7   7   4   5  32  26  15   5   4   
10      1  14  15  13  13  16  15  17   8  11  11   5   8  19  30  20   9   5   
11      1  19  16  12  22  12  18  16   7  17   6   7  10  27  19  10  10   4   
12      5  34  12  24  20  27  23  34  13  43  23   7  21  30  35  27  14  13   

topic  18  19  20  21  22  23  24  25  26   27  28  29  30  31  32  33  34  \
month                                                                        
1       6  25  11  34  37  23  40   6  31   53   9  19  33  17   8  48  26   
2       5  31  13  19  20  26  30  10  16   73   6  19  26  15  12  39  19   
3       5  41  14  35  39  26  39  13  51   88   4  25  25  18  14  74  24   
4       9  39  24  28  31  28  40  18  37  108  12  22  22  21  20  57  18   
5       9  41  16  22  38  20  37  18  45   99  15  17  16  22  12  58  23   
6       7  34  15  33  34  16  35  16  41   71  12  22  23  19   9  46  25   
7      14  29  15  32  34  19  26  18  38   74   8  25  21  15  11  54  17   
8       3  16   4  15  12  14  14   4  23   60   2   3  15  13   8  28  14   
9       1  14  12  13  29   8  15   7  14   60   5  18   9  12   4  30  13   
10      3  17   6  23  16   6  17   4  18   50   5   9  15  12   4  13  11   
11      5  17   7  17  16  10  22   3  18   43   6   9  13  12   5  32  15   
12      7  32  15  21  29  24  25  10  29   72  11  18  36  10  11  45  16   

topic  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  \
month                                                                           
1      35  36  81  19  17  11  25  45  21  25  19  23  20  55  29  17  31   9   
2      35  29  63  20   8   8  31  38  21  15  15  22  26  58  30  17  35   8   
3      43  26  75  31  23  12  25  62  20  19  21  35  39  64  28  28  45  16   
4      35  33  73  34  24  16  46  55  22  24  19  37  27  57  40  17  44  26   
5      36  28  53  39  20  16  35  64  20  18  21  37  42  45  33  11  35  12   
6      49  27  64  26  13   8  40  57  19  15  16  39  46  43  29  19  37  13   
7      44  32  66  31  20  15  30  74  18  18  18  25  34  46  43  11  35  15   
8      21  15  51  19  10   5  18  18  14  10  11  18  22  29  21  10  15   9   
9      20  13  51  20  12   6  25  18   8   7  10  20  15  29  14   7  21   6   
10     16  16  47  17   8   4  17  26   5  12  11  17  21  22  20  10  20   4   
11     16  10  53  15   7   5  14  32  13  12   8  24  20  19  13  13  18   4   
12     32  32  51  25  20   9  25  33  19  15  19  26  22  42  37  15  31   5   

topic  53  54  55  56  57  58  59   60  61  62  63  64  65  66   67  68  69  \
month                                                                         
1       6  12  25  45  15  21  20  148  19  16  13  28  42  12   57  21   7   
2      11  13  20  31  11  15  23  111  17  13  11  21  37   9   47  25  12   
3       9  12  20  33  30  16  22  190  20  25  10  21  40  20   93  38  22   
4       7  10  13  34  16  25  33  157  23  19  13  22  41  10   67  53  14   
5       8  13  13  45  16  17  27  172  23  21   8  35  25  16   56  31  19   
6       8  12  16  37  21  16  18  213  28  26  15  37  32  27   92  31  20   
7       6   9  13  19  24  18  22  162  30  26  15  30  33  46  197  30  18   
8       7   9  11  26  11  10  16   56   9  10   6  12  12   7   29  14  11   
9       3   8  12  30  10  12  11   97  13  10   5  17  23   7   35  22  12   
10      8   6   6  27   5   7  14   88  13  11   8   6  33   5   30  17  10   
11      4   9  16  26  11  10   8   61  10  12   8  14  39   7   42  17   7   
12      6  14  32  30  13  21  29  116  28  25   8  29  24  15   62  30  11   

topic  70  71  72  73  74  
month                      
1      25  25   2  38  24  
2      15  26   4  32  18  
3      27  31   7  30  29  
4      34  33   2  29  35  
5      11  22   3  27  39  
6      18  19   4  31  41  
7      17  22   4  31  23  
8      14  12   7  14  11  
9       8  13   0  20  14  
10     11   9   1  15  13  
11     13  17   0  23  13  
12     24  32   2  38  16

import numpy as np
import pandas as pd

# Ensure reviewTime is datetime
reviews_df['reviewTime'] = pd.to_datetime(reviews_df['reviewTime'])
reviews_df['month'] = reviews_df['reviewTime'].dt.month

# Count reviews per month per topic
seasonal_topics = reviews_df.groupby(['month', 'topic']).size().unstack(fill_value=0)

# Log-transform counts (add 1 to avoid log(0))
log_counts = np.log1p(seasonal_topics)

# Mean across topics for each month
month_popularity = log_counts.sum(axis=1)

# Display
print("Log-transformed summed review counts per month:")
print(month_popularity)

Log-transformed summed review counts per month:
month
1     234.608708
2     226.930932
3     246.091147
4     247.316282
5     240.029166
6     242.125116
7     242.570914
8     192.285163
9     191.662491
10    189.558463
11    193.861213
12    231.726066
dtype: float64

import plotly.graph_objs as go

# month_popularity from last snippet (total reviews per month)
fig = go.Figure(
    data=go.Bar(
        x=month_popularity.index, 
        y=month_popularity.values,
        text=month_popularity.values,  # show raw counts on hover
        textposition='auto',
        marker_color='teal',
        hovertemplate="Month: %{x}<br>Total Reviews: %{y}<extra></extra>"
    )
)

fig.update_layout(
    title="Total Reviews per Month Across All Topics (log transformed)",
    xaxis_title="Month",
    yaxis_title="Total Reviews",
    xaxis=dict(tickmode='linear'),
    yaxis=dict(range=[150, 250]),  # zoomed range
    template='plotly_white',
    height=500
)

fig.show()

# 4. SENTIMENT ANALYSIS ENHANCEMENT 
# ===============================================
print("\n4. Enhanced Sentiment Analysis")
print("="*40)

from textblob import TextBlob

# Add sentiment scores
def get_sentiment(text):
    return TextBlob(str(text)).sentiment.polarity

reviews_df['sentiment_score'] = reviews_df['reviewText'].apply(get_sentiment)
reviews_df['sentiment_category'] = pd.cut(
    reviews_df['sentiment_score'], 
    bins=[-1, -0.1, 0.1, 1], 
    labels=['Negative', 'Neutral', 'Positive']
)

# Analyze topics by sentiment
sentiment_topics = reviews_df.groupby(['sentiment_category', 'topic']).size().unstack(fill_value=0)
print("Topics by sentiment category:")
print(sentiment_topics)

4. Enhanced Sentiment Analysis
========================================
Topics by sentiment category:
topic               0    1    2    3    4    5    6    7    8    9    10  11  \
sentiment_category                                                             
Negative             3   18    4   14   11   10   13   11    8    3    8  26   
Neutral             34   42   28   22   41   34   50   58   37  201   23  29   
Positive            44  242  215  192  231  239  211  343  148   44  140  78   

topic                12   13   14   15   16  17  18   19   20   21   22   23  \
sentiment_category                                                             
Negative              4   16   29   13    8   5   3    7    1    6   11    5   
Neutral              25   72   70   43   22  14  18   42   12   25   42   17   
Positive            179  295  347  279  121  82  53  287  139  261  282  198   

topic                24  25   26   27  28   29   30   31  32   33   34   35  \
sentiment_category                                                            
Negative              9  13   17   35   7   12   14   10   2   31   10    9   
Neutral              53  18   38  222  17   35   45   33  18   85   29   28   
Positive            278  96  306  594  71  159  195  143  98  408  182  345   

topic                36   37   38   39  40   41   42   43   44   45   46   47  \
sentiment_category                                                              
Negative              6   41   11   17   5   15    5   18   18    3   18   25   
Neutral              20  187   64   29  27   62   30   43   43   23   70   86   
Positive            271  500  221  136  83  254  487  139  129  162  235  223   

topic                48   49   50   51  52  53   54   55   56   57   58   59  \
sentiment_category                                                             
Negative              5   11    4   11  11   3    2    3   20    8   15   33   
Neutral              79   32   24   38  17  23   12   17   70   23   37   61   
Positive            425  294  147  318  99  57  113  177  293  152  134  148   

topic                60   61   62  63   64   65   66   67   68   69   70   71  \
sentiment_category                                                              
Negative            120   10    4   6    9   10   10    6   15   16   13   14   
Neutral             487   41   17  17   29   78   17   41   66   40   27   27   
Positive            964  182  193  97  234  293  154  760  248  107  177  220   

topic               72   73   74  
sentiment_category                
Negative             0   10    6  
Neutral              6   48   33  
Positive            30  270  237

# 4b. TOP NEGATIVELY PERCEIVED TOPICS BY NEG/POS RATIO
# =====================================================
print("\nTop Negatively Perceived Topics (by Negative/Positive Ratio)")
print("="*60)

# Aggregate counts of each sentiment per topic (if not already)
sentiment_topics = reviews_df.groupby(['topic', 'sentiment_category']).size().unstack(fill_value=0)

# Compute Negative/Positive ratio
sentiment_topics['neg_pos_ratio'] = sentiment_topics.apply(
    lambda row: row['Negative'] / row['Positive'] if row['Positive'] > 0 else float('inf'),
    axis=1
)

# Sort topics by negative-to-positive ratio
top_negative_topics = sentiment_topics.sort_values('neg_pos_ratio', ascending=False)

# Display top 10 most negatively perceived topics
print(top_negative_topics[['Negative', 'Positive', 'Neutral', 'neg_pos_ratio']].head(10))

Top Negatively Perceived Topics (by Negative/Positive Ratio)
============================================================
sentiment_category  Negative  Positive  Neutral  neg_pos_ratio
topic                                                         
11                        26        78       29       0.333333
59                        33       148       61       0.222973
69                        16       107       40       0.149533
44                        18       129       43       0.139535
25                        13        96       18       0.135417
43                        18       139       43       0.129496
39                        17       136       29       0.125000
60                       120       964      487       0.124481
47                        25       223       86       0.112108
58                        15       134       37       0.111940

# 5. PRICE SENSITIVITY ANALYSIS (Intermediate)
# ============================================
print("\n5. Price Sensitivity Analysis")
print("="*40)

# Look for price-related keywords in reviews
price_keywords = ['price', 'expensive', 'cheap', 'value', 'worth', 'cost', 'deal', 'money']
price_mentions = reviews_df[reviews_df['reviewText'].str.contains('|'.join(price_keywords), case=False, na=False)]

print(f"Reviews mentioning price: {len(price_mentions)} out of {len(reviews_df)} ({len(price_mentions)/len(reviews_df)*100:.1f}%)")

# Analyze topics in price-sensitive reviews
price_topics = price_mentions['topic'].value_counts().head(10)
print("\nTop topics in price-sensitive reviews:")
print(price_topics)

5. Price Sensitivity Analysis
========================================
Reviews mentioning price: 3662 out of 21570 (17.0%)

Top topics in price-sensitive reviews:
topic
37    175
27    175
33    175
60    153
67    115
14     95
48     82
47     80
35     78
13     75
Name: count, dtype: int64

# 6. PRODUCT LIFECYCLE ANALYSIS (Advanced)
# ========================================
print("\n6. Product Lifecycle Analysis")
print("="*40)

# Analyze how topics change for products over time since launch
product_lifecycle = reviews_df.groupby(['asin', 'year', 'topic']).size().reset_index(name='count')

# Find products with long review histories
products_with_history = reviews_df.groupby('asin')['year'].nunique()
long_history_products = products_with_history[products_with_history >= 3]

print(f"Products with 3+ years of reviews: {len(long_history_products)}")

print(long_history_products.sort_values(ascending=False).head(10))

6. Product Lifecycle Analysis
========================================
Products with 3+ years of reviews: 365
asin
B000E8J5PY    7
B000A8BUEC    6
B000NOCJJS    6
B001V6PNRW    6
B000A8BU0Q    6
B0030NKIH4    5
B000E8F2TM    5
B000E8H01K    5
B000E8H0LA    5
B003QL2SQO    5
Name: year, dtype: int64

# 8. REVIEW DEPTH ANALYSIS (Intermediate)
# =======================================
print("\n8. Review Depth Analysis")
print("="*40)

# Analyze review length and its relationship to topics
reviews_df['review_length'] = reviews_df['reviewText'].str.len()
reviews_df['review_depth'] = pd.cut(
    reviews_df['review_length'], 
    bins=[0, 100, 500, float('inf')], 
    labels=['Short', 'Medium', 'Long']
)

depth_topics = reviews_df.groupby(['review_depth', 'topic']).size().unstack(fill_value=0)
print("Topics by review depth:")
print(depth_topics.head())

8. Review Depth Analysis
========================================
Topics by review depth:
topic         0    1    2    3    4    5    6    7    8    9    10   11   12  \
review_depth                                                                   
Short         40   12   15   15    9   15   17   40   17   22   25   12   41   
Medium        41  273  217  208  243  260  238  353  172  222  146  115  163   
Long           0   17   15    5   32    8   19   20    4    4    0    6    4   

topic          13   14   15   16  17  18   19   20   21   22   23   24   25  \
review_depth                                                                  
Short          14   40   35   11  11   9   27   14   29   27   24   14   20   
Medium        278  382  270  136  89  63  295  138  259  299  196  306  104   
Long           91   24   31    4   1   2   14    0    4    9    0   20    3   

topic          26   27  28   29   30   31   32   33   34   35   36   37   38  \
review_depth                                                                   
Short          23   12  10   17   17   10   11   17   14   36   35   22   15   
Medium        318  499  79  183  227  171  103  426  193  342  259  510  260   
Long           20  340   6    6   10    5    4   81   14    4    3  196   21   

topic          39  40   41   42   43   44   45   46   47   48   49   50   51  \
review_depth                                                                   
Short          14  13    9   55   16    8   15   19   14   47   38   10   36   
Medium        163  99  260  445  176  172  168  274  265  415  292  149  325   
Long            5   3   62   22    8   10    5   30   55   47    7   16    6   

topic          52  53   54   55   56   57   58   59    60   61   62  63   64  \
review_depth                                                                   
Short          10  10   11   17    9   23   10   16   129   15   25  21   11   
Medium        113  73  111  179  302  158  170  203  1303  202  185  96  253   
Long            4   0    5    1   72    2    8   24   139   16    4   3    8   

topic          65   66   67   68   69   70   71  72   73   74  
review_depth                                                   
Short          23   46  215   11   12   12   10   1   21   12  
Medium        298  127  588  266  146  193  240  33  291  249  
Long           60    8    4   52    5   12   11   2   16   15

# 9. FEATURE IMPORTANCE ANALYSIS (Advanced)
# =========================================
print("\n9. Feature Importance Analysis")
print("="*40)

# Extract specific product features mentioned
feature_keywords = {
    'comfort': ['comfort', 'comfortable', 'cozy'],
    'durability': ['durable', 'lasting', 'sturdy', 'quality'],
    'design': ['design', 'style', 'look', 'appearance'],
    'fit': ['fit', 'size', 'sizing', 'tight', 'loose'],
    'performance': ['performance', 'speed', 'grip', 'traction']
}

for feature, keywords in feature_keywords.items():
    feature_reviews = reviews_df[reviews_df['reviewText'].str.contains('|'.join(keywords), case=False, na=False)]
    feature_topics = feature_reviews['topic'].value_counts().head(5)
    print(f"\nTop topics mentioning {feature}:")
    print(feature_topics)

9. Feature Importance Analysis
========================================

Top topics mentioning comfort:
topic
60    386
27    361
67    254
42    221
14    170
Name: count, dtype: int64

Top topics mentioning durability:
topic
27    116
60    111
48    110
51     77
42     71
Name: count, dtype: int64

Top topics mentioning design:
topic
60    388
27    291
37    249
67    216
42    187
Name: count, dtype: int64

Top topics mentioning fit:
topic
60    1338
67     395
27     369
33     212
42     205
Name: count, dtype: int64

Top topics mentioning performance:
topic
27    59
71    32
37    31
61    26
60    26
Name: count, dtype: int64

# 10. ACTIONABLE INSIGHTS GENERATOR 
# ============================================
print("\n10. Actionable Insights Generator")
print("="*40)

from collections import Counter
import re

def get_top_words(text_series, n_words=10):
    """Extract top words from a series of text"""
    # Combine all text and clean it
    all_text = ' '.join(text_series.astype(str))
    # Remove punctuation and convert to lowercase
    words = re.findall(r'\b[a-zA-Z]{3,}\b', all_text.lower())
    # Remove common stopwords
    stopwords = {'the', 'and', 'for', 'are', 'but', 'not', 'you', 'all', 'can', 'had', 'her', 'was', 'one', 'our', 'out', 'day', 'get', 'has', 'him', 'his', 'how', 'man', 'new', 'now', 'old', 'see', 'two', 'way', 'who', 'boy', 'did', 'its', 'let', 'put', 'say', 'she', 'too', 'use', 'this', 'that', 'with', 'have', 'from', 'they', 'know', 'want', 'been', 'good', 'much', 'some', 'time', 'very', 'when', 'come', 'here', 'just', 'like', 'long', 'make', 'many', 'over', 'such', 'take', 'than', 'them', 'well', 'were', 'will'}
    filtered_words = [word for word in words if word not in stopwords and len(word) > 3]
    return Counter(filtered_words).most_common(n_words)

def generate_insights(topic_id):
    """Generate specific insights for a topic"""
    topic_reviews = reviews_df[reviews_df['topic'] == topic_id]
    
    insights = {
        'topic_id': topic_id,
        'total_reviews': len(topic_reviews),
        'avg_rating': topic_reviews['overall'].mean(),
        'sentiment_distribution': topic_reviews['sentiment_category'].value_counts() if 'sentiment_category' in reviews_df.columns else 'N/A',
        'top_products': topic_reviews['asin'].value_counts().head(3),
        'seasonal_pattern': topic_reviews['month'].value_counts().head(3) if 'month' in reviews_df.columns else 'N/A',
        'top_words': get_top_words(topic_reviews['reviewText'], 15),
        'sample_reviews': topic_reviews['reviewText'].head(3).tolist()
    }
    
    return insights

# Example: Generate insights for topics
top_topics = reviews_df['topic'].value_counts().head(30).index
for topic in top_topics:
    insights = generate_insights(topic)
    print(f"\nInsights for Topic {topic}:")
    print(f"- Reviews: {insights['total_reviews']}")
    print(f"- Avg Rating: {insights['avg_rating']:.2f}")
    print(f"- Top Product: {insights['top_products'].index[0]} ({insights['top_products'].iloc[0]} reviews)")
    print(f"- Top Words: {', '.join([f'{word}({count})' for word, count in insights['top_words'][:10]])}")
    print(f"- Product Clues: {', '.join([word for word, count in insights['top_words'][:15] if any(keyword in word.lower() for keyword in ['shoe', 'boot', 'sneaker', 'shirt', 'jacket', 'watch', 'phone', 'case', 'bag', 'nike', 'adidas', 'apple', 'samsung'])])}")

10. Actionable Insights Generator
========================================

Insights for Topic 60:
- Reviews: 1571
- Avg Rating: 3.86
- Top Product: B000V4YZ1K (46 reviews)
- Top Words: size(1847), shoes(984), these(956), shoe(658), small(574), nike(428), wear(415), half(413), would(387), great(382)
- Product Clues: shoes, shoe, nike

Insights for Topic 27:
- Reviews: 851
- Avg Rating: 3.98
- Top Product: B004IM1GHW (35 reviews)
- Top Words: shoes(1235), these(1138), shoe(861), nike(609), feet(392), running(391), pair(390), comfortable(337), wear(317), great(313)
- Product Clues: shoes, shoe, nike

Insights for Topic 67:
- Reviews: 807
- Avg Rating: 4.79
- Top Product: B000V4YZ1K (15 reviews)
- Top Words: shoes(338), these(329), great(301), comfortable(245), love(219), perfect(161), exactly(158), shoe(126), color(124), what(121)
- Product Clues: shoes, shoe

Insights for Topic 37:
- Reviews: 728
- Avg Rating: 3.70
- Top Product: B007FXKMLW (13 reviews)
- Top Words: watch(1019), nike(335), these(263), great(251), shoes(209), band(208), would(196), only(195), battery(163), bought(153)
- Product Clues: watch, nike, shoes, shoe

Insights for Topic 33:
- Reviews: 524
- Avg Rating: 4.17
- Top Product: B004IM1GHW (13 reviews)
- Top Words: these(490), shoes(474), shoe(266), nike(187), great(178), comfortable(157), pair(155), feet(150), wear(127), running(107)
- Product Clues: shoes, shoe, nike

Insights for Topic 42:
- Reviews: 522
- Avg Rating: 4.72
- Top Product: B0098SW6BG (8 reviews)
- Top Words: these(372), shoes(324), comfortable(220), great(216), love(154), shoe(143), wear(106), size(106), nike(101), awesome(93)
- Product Clues: shoes, shoe, nike

Insights for Topic 48:
- Reviews: 509
- Avg Rating: 4.51
- Top Product: B00515GDJG (89 reviews)
- Top Words: great(175), shoes(162), perfect(120), size(120), clothes(101), love(90), these(89), would(87), carry(80), small(73)
- Product Clues: shoes, nike, shoe

Insights for Topic 14:
- Reviews: 446
- Avg Rating: 4.23
- Top Product: B004IM1GHW (26 reviews)
- Top Words: shoes(302), these(292), comfortable(174), shoe(158), pair(139), ever(128), nike(125), great(121), best(105), love(90)
- Product Clues: shoes, shoe, nike

Insights for Topic 7:
- Reviews: 413
- Avg Rating: 4.56
- Top Product: B007FXKMLW (13 reviews)
- Top Words: shoes(216), these(215), great(137), comfortable(91), pair(76), shoe(76), would(71), bought(69), nike(68), thank(64)
- Product Clues: shoes, shoe, nike

Insights for Topic 13:
- Reviews: 383
- Avg Rating: 4.13
- Top Product: B0098G7Q1S (8 reviews)
- Top Words: shoes(356), these(355), shoe(199), nike(152), great(139), comfortable(134), feet(130), running(114), pair(111), wear(93)
- Product Clues: shoes, shoe, nike

Insights for Topic 56:
- Reviews: 383
- Avg Rating: 3.92
- Top Product: B000E8H01K (6 reviews)
- Top Words: watch(403), these(164), nike(164), great(160), shoes(127), would(82), size(80), face(78), pair(78), shoe(74)
- Product Clues: watch, nike, shoes, shoe

Insights for Topic 35:
- Reviews: 382
- Avg Rating: 4.62
- Top Product: B004IM1GHW (8 reviews)
- Top Words: shoes(211), great(192), these(184), recommend(139), comfortable(135), shoe(120), would(107), anyone(83), love(77), nike(61)
- Product Clues: shoes, shoe, nike

Insights for Topic 65:
- Reviews: 381
- Avg Rating: 4.12
- Top Product: B007FXKMLW (13 reviews)
- Top Words: watch(415), great(136), these(118), nike(102), wrist(100), shoes(98), comfortable(97), easy(83), band(81), shoe(75)
- Product Clues: watch, nike, shoes, shoe

Insights for Topic 51:
- Reviews: 367
- Avg Rating: 4.61
- Top Product: B000V4YZ1K (7 reviews)
- Top Words: shoes(182), these(146), great(105), comfortable(86), nike(76), love(71), nice(70), quality(67), shoe(67), look(64)
- Product Clues: shoes, nike, shoe

Insights for Topic 26:
- Reviews: 361
- Avg Rating: 4.54
- Top Product: B002JB6VEC (17 reviews)
- Top Words: these(252), shoes(218), comfortable(163), shoe(121), pair(118), great(98), feet(95), work(90), nike(88), bought(74)
- Product Clues: shoes, shoe, nike

Insights for Topic 24:
- Reviews: 340
- Avg Rating: 4.42
- Top Product: B001V6PNRW (8 reviews)
- Top Words: shoes(218), these(196), great(101), shoe(91), comfortable(87), size(72), nike(67), love(64), would(54), pair(48)
- Product Clues: shoes, shoe, nike

Insights for Topic 49:
- Reviews: 337
- Avg Rating: 4.61
- Top Product: B003NYUSPM (12 reviews)
- Top Words: these(204), shoes(166), great(112), comfortable(108), shoe(95), excellent(81), socks(59), bought(54), nike(53), loves(48)
- Product Clues: shoes, shoe, nike

Insights for Topic 19:
- Reviews: 336
- Avg Rating: 4.51
- Top Product: B007FXKMLW (5 reviews)
- Top Words: shoes(186), these(172), shoe(107), great(91), comfortable(72), pair(67), love(65), nike(58), bought(48), size(46)
- Product Clues: shoes, shoe, nike

Insights for Topic 15:
- Reviews: 336
- Avg Rating: 4.26
- Top Product: B007FXKMLW (8 reviews)
- Top Words: shoes(280), these(256), shoe(170), great(118), comfortable(92), nike(86), running(85), support(85), pair(80), feet(79)
- Product Clues: shoes, shoe, nike

Insights for Topic 22:
- Reviews: 335
- Avg Rating: 4.40
- Top Product: B0006NGUE6 (9 reviews)
- Top Words: shoes(227), these(214), comfortable(146), great(117), shoe(107), golf(83), wear(75), love(70), look(69), size(58)
- Product Clues: shoes, shoe, nike

Insights for Topic 47:
- Reviews: 334
- Avg Rating: 3.71
- Top Product: B004IM1GHW (20 reviews)
- Top Words: shoes(258), these(252), shoe(169), nike(149), comfortable(114), size(86), great(86), pair(81), wear(69), more(60)
- Product Clues: shoes, shoe, nike

Insights for Topic 41:
- Reviews: 331
- Avg Rating: 4.19
- Top Product: B006O37RY4 (10 reviews)
- Top Words: shoes(343), these(307), nike(207), shoe(187), pair(137), running(127), comfortable(118), great(118), feet(82), vomero(74)
- Product Clues: shoes, nike, shoe

Insights for Topic 68:
- Reviews: 329
- Avg Rating: 4.24
- Top Product: B003NYUSPM (14 reviews)
- Top Words: these(267), shoes(244), shoe(159), great(122), nike(106), comfortable(100), feet(89), socks(71), pair(66), size(63)
- Product Clues: shoes, shoe, nike

Insights for Topic 73:
- Reviews: 328
- Avg Rating: 4.34
- Top Product: B003NYUSPM (27 reviews)
- Top Words: these(228), socks(128), great(117), shoes(112), feet(109), comfortable(99), shoe(82), wear(72), nike(63), love(57)
- Product Clues: shoes, shoe, nike, boots

Insights for Topic 46:
- Reviews: 323
- Avg Rating: 4.02
- Top Product: B000V4YZ1K (18 reviews)
- Top Words: these(240), shoes(193), shoe(113), great(103), comfortable(99), pair(73), feet(70), little(69), size(66), nike(65)
- Product Clues: shoes, shoe, nike

Insights for Topic 1:
- Reviews: 302
- Avg Rating: 4.32
- Top Product: B000V4YZ1K (11 reviews)
- Top Words: these(189), shoes(166), comfortable(100), shoe(97), nike(78), pair(78), great(75), feet(61), size(61), bought(58)
- Product Clues: shoes, shoe, nike

Insights for Topic 36:
- Reviews: 297
- Avg Rating: 4.72
- Top Product: B0036DRWKO (18 reviews)
- Top Words: shoes(128), these(125), great(82), loves(64), comfortable(64), grandson(63), thanks(60), love(50), nike(46), happy(46)
- Product Clues: shoes, nike, shoe

Insights for Topic 38:
- Reviews: 296
- Avg Rating: 4.26
- Top Product: B000V4YZ1K (6 reviews)
- Top Words: shoes(181), these(148), shoe(119), size(96), great(96), comfortable(90), nike(74), wide(67), love(67), width(64)
- Product Clues: shoes, shoe, nike

Insights for Topic 21:
- Reviews: 292
- Avg Rating: 4.65
- Top Product: B004UBXO9Y (6 reviews)
- Top Words: shoes(154), great(106), these(98), love(76), comfortable(66), confortable(60), shoe(59), pair(53), nice(48), nike(48)
- Product Clues: shoes, shoe, nike

Insights for Topic 4:
- Reviews: 284
- Avg Rating: 4.34
- Top Product: B004IM1GHW (5 reviews)
- Top Words: shoes(228), these(215), shoe(121), comfortable(108), nike(101), great(90), pair(82), wear(67), feet(58), running(56)
- Product Clues: shoes, shoe, nike

import pandas as pd
import matplotlib.pyplot as plt

uniquereviewers = reviews_df['reviewerID'].nunique()
print(f'Number of unique reviewers: {uniquereviewers}')  # Total unique reviewers in the dataset

# --- Top Reviewers Script ---
def plot_top_reviewers(df, n=10):
    """
    Show top reviewers by number of reviews.
    Args:
        df: DataFrame with 'reviewerID' column
        n: number of top reviewers to display
    """
    # Count reviews per reviewer
    top_reviewers = df['reviewerID'].value_counts().head(n)

    # Print table for reference

    print("Top Reviewers:")
    print(top_reviewers)
    
    
    # Plot
    plt.figure(figsize=(10, 6))
    top_reviewers.plot(kind='bar', color='steelblue', edgecolor='black')
    plt.title(f"Top {n} Reviewers by Number of Reviews")
    plt.xlabel("Reviewer ID")
    plt.ylabel("Number of Reviews")
    plt.xticks(rotation=45, ha='right')
    plt.tight_layout()
    plt.show()

# --- Example usage ---
plot_top_reviewers(reviews_df, n=10)

Number of unique reviewers: 20423
Top Reviewers:
reviewerID
A32D0TORRTI9E5    11
A265PCQZA1Y1WF     7
A237P6ACWJR9LM     7
ARE2MDLUM3APW      7
A1PR0WW5UQHL7X     6
A3HJOKYEV152QJ     5
A1OTV3X29IALOU     5
A3FP6ZIBTMCPCF     5
AC441KJETNFXS      5
A2J3QP96I96P4B     5
Name: count, dtype: int64

import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns

def plot_top_reviewer_topics(df, n=10):
    """
    Show distribution of topics reviewed by top reviewers.
    Args:
        df: DataFrame with 'reviewerID' and 'topic' columns
        n: number of top reviewers to display
    """
    # Get top reviewers
    top_reviewers = df['reviewerID'].value_counts().head(n).index

    # Filter down to their reviews
    subset = df[df['reviewerID'].isin(top_reviewers)]

    # Build pivot table: reviewers (rows) x topics (columns)
    reviewer_topics = pd.crosstab(subset['reviewerID'], subset['topic'])

    # Print table for reference
    print("Top Reviewer → Topic counts:")
    print(reviewer_topics)

# --- Example usage ---
plot_top_reviewer_topics(reviews_df, n=10)

Top Reviewer → Topic counts:
topic           0   2   3   4   5   6   7   8   10  12  14  19  22  23  26  \
reviewerID                                                                   
A1OTV3X29IALOU   0   0   1   0   0   0   0   0   0   0   1   0   0   0   1   
A1PR0WW5UQHL7X   0   0   0   0   1   0   0   0   1   0   0   0   0   1   0   
A237P6ACWJR9LM   0   0   0   0   1   1   0   0   0   0   0   0   1   0   0   
A265PCQZA1Y1WF   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   
A2J3QP96I96P4B   0   0   0   0   0   0   0   0   0   0   0   0   0   0   0   
A32D0TORRTI9E5   0   3   0   0   0   0   0   0   0   0   0   0   1   0   0   
A3FP6ZIBTMCPCF   1   0   0   0   0   1   0   0   0   0   0   0   0   0   0   
A3HJOKYEV152QJ   0   0   0   1   0   0   0   1   0   0   1   1   0   0   0   
AC441KJETNFXS    0   1   0   0   0   0   1   0   0   0   0   0   0   0   0   
ARE2MDLUM3APW    0   0   0   0   0   0   0   0   0   2   0   0   0   0   0   

topic           27  31  33  35  36  37  38  41  42  43  47  51  56  60  61  \
reviewerID                                                                   
A1OTV3X29IALOU   0   0   0   0   0   0   0   0   1   1   0   0   0   0   0   
A1PR0WW5UQHL7X   0   0   0   0   0   0   1   0   0   0   0   0   0   1   0   
A237P6ACWJR9LM   0   0   0   0   0   1   0   1   0   0   1   0   0   0   0   
A265PCQZA1Y1WF   1   0   0   0   0   0   0   4   0   0   0   0   1   0   1   
A2J3QP96I96P4B   1   0   0   1   0   0   0   0   2   0   0   0   0   0   0   
A32D0TORRTI9E5   0   0   1   0   0   0   1   0   1   0   0   1   1   0   0   
A3FP6ZIBTMCPCF   0   0   0   0   0   0   0   0   1   1   0   0   0   1   0   
A3HJOKYEV152QJ   1   0   0   0   0   0   0   0   0   0   0   0   0   0   0   
AC441KJETNFXS    0   0   0   0   1   0   0   0   0   0   0   0   0   1   0   
ARE2MDLUM3APW    0   1   0   2   0   0   0   0   2   0   0   0   0   0   0   

topic           64  67  73  74  
reviewerID                      
A1OTV3X29IALOU   0   0   0   0  
A1PR0WW5UQHL7X   1   0   0   0  
A237P6ACWJR9LM   0   0   0   1  
A265PCQZA1Y1WF   0   0   0   0  
A2J3QP96I96P4B   0   0   1   0  
A32D0TORRTI9E5   2   0   0   0  
A3FP6ZIBTMCPCF   0   0   0   0  
A3HJOKYEV152QJ   0   0   0   0  
AC441KJETNFXS    0   1   0   0  
ARE2MDLUM3APW    0   0   0   0

Topic Modeling Nike Product Reviews: Marketing and Product Insights¶

Introduction¶

Methods¶

Data Cleaning¶

Vectorization¶

Topic Modeling¶

Sentiment Analysis¶

Results¶

Topic Distribution¶

Reviewer Patterns¶

Seasonality Trends¶

Sentiment by Topic¶

Product-Level Insights¶

Conclusion¶

Future Work:¶

13 topics¶

29 topics¶

75 topics (most definitive)¶

	asin	reviewText	overall	reviewerID	reviewTime	cleaned
0	B0000V9K32	the colour i received is not blue as shown but...	3.0	ACT5DY536GISV	07 27, 2009	colour received blue shown yellowcouldnt chang...
1	B0000V9K32	Very cute and is really practical. Fits better...	5.0	A3BVWMS9I8OH8U	12 26, 2009	cute really practical fit better smaller wrist...
2	B0000V9K3W	The watch was exactly what i ordered and I got...	4.0	A5RZS69KSJH00	10 24, 2009	watch exactly ordered got fast unfortunately b...
3	B0000V9K46	This product came promptly and as described, p...	5.0	A3F8O512N9UNVM	03 26, 2008	came promptly described pleasure business themd
4	B0000V9KNM	Why isn't Nike making these anymore? I love t...	4.0	A2EAKTCKFRF7A4	11 16, 2011	isnt making anymore love watch get lot complim...
...	...	...	...	...	...	...
21565	B00KRENKNS	love it	5.0	AQOZEPFW4HTAR	07 15, 2014	love
21566	B00KXBJQ1A	They fit perfect and the are really light	5.0	A3H37LNK1EFP7C	07 21, 2014	fit perfect really light
21567	B00L01ML5K	The product was shipped in record speed! I lov...	5.0	A1N0CN53K1G2RW	07 6, 2014	shipped record speed love sneaker look feel aw...
21568	B00L2ADWCA	I like this item but they didn't have my size ...	5.0	AP318RPSIQQFA	07 1, 2014	like item didnt size tild alert
21569	B00L5K86LO	Love it !	5.0	A1KBC812A7RSY9	07 20, 2014	love

	reviewText	topic	topic_top_words
0	the colour i received is not blue as shown but...	64	gotten, many, strong, shown, compliment, surpr...
1	Very cute and is really practical. Fits better...	19	room, sister, backpack, school, penny, flexibl...
2	The watch was exactly what i ordered and I got...	59	ask, refund, xmas, sent, attention, get, received
3	This product came promptly and as described, p...	17	advertised, rating, cold, chose, promptly, wea...
4	Why isn't Nike making these anymore? I love t...	74	pant, fit, smooth, great, love, ride, good

	topic	asin	count
0	48	B00515GDJG	89
1	48	B00515GEJK	61
2	60	B000V4YZ1K	46
3	60	B003HD6KY2	38
4	27	B004IM1GHW	35
5	60	B0098G7Q1S	32
6	48	B004YI0X6A	31
7	73	B003NYUSPM	27
8	14	B004IM1GHW	26
9	27	B007FXKMLW	21
10	47	B004IM1GHW	20
11	46	B000V4YZ1K	18
12	36	B0036DRWKO	18
13	73	B000V4YZ1K	18
14	26	B002JB6VEC	17
15	67	B000V4YZ1K	15
16	68	B003NYUSPM	14
17	67	B006RBH2PW	13
18	65	B007FXKMLW	13
19	33	B004IM1GHW	13
20	37	B007FXKMLW	13
21	7	B007FXKMLW	13
22	33	B007FXKMLW	12
23	67	B000G42Z2Q	12
24	49	B003NYUSPM	12

	topic	asin	review_count
0	0	B00AMTMRQE	2
1	1	B000V4YZ1K	11
2	2	B007FXKMLW	6
3	3	B0098G7Q1S	4
4	4	B007FXKMLW	5
...	...	...	...
70	70	B000V4YZ1K	4
71	71	B001LDDR0K	6
72	72	B005OBFUU0	2
73	73	B003NYUSPM	27
74	74	B003NYUSPM	4