【Python】商品レビューの分析: 顧客の声を活用する方法

import requests
from bs4 import BeautifulSoup

# 例: 商品ページからレビューを収集する
url = 'https://example.com/product_reviews'
response = requests.get(url)
soup = BeautifulSoup(response.text, 'html.parser')

reviews = soup.find_all('div', class_='review')
review_texts = [review.get_text() for review in reviews]

import re

def clean_text(text):
    text = re.sub(r'<.*?>', '', text)  # HTMLタグを削除
    text = re.sub(r'\d+', '', text)  # 数字を削除
    text = text.lower().strip()  # 小文字に変換し、余分な空白を削除
    return text

cleaned_reviews = [clean_text(review) for review in review_texts]

from textblob import TextBlob

def get_sentiment(text):
    blob = TextBlob(text)
    sentiment = blob.sentiment.polarity  # ポジティブな値はポジティブ、ネガティブな値はネガティブ
    return sentiment

sentiments = [get_sentiment(review) for review in cleaned_reviews]

from sklearn.feature_extraction.text import CountVectorizer
from sklearn.decomposition import LatentDirichletAllocation

vectorizer = CountVectorizer(stop_words='english')
X = vectorizer.fit_transform(cleaned_reviews)

lda = LatentDirichletAllocation(n_components=5, random_state=0)
lda.fit(X)

# トピックの表示
for topic_idx, topic in enumerate(lda.components_):
    print(f"Topic {topic_idx}:")
    print([vectorizer.get_feature_names_out()[i] for i in topic.argsort()[:-10 - 1:-1]])

import matplotlib.pyplot as plt
from wordcloud import WordCloud

# ポジティブ・ネガティブの割合
positive_reviews = sum(1 for sentiment in sentiments if sentiment > 0)
negative_reviews = sum(1 for sentiment in sentiments if sentiment < 0)

labels = ['Positive', 'Negative']
sizes = [positive_reviews, negative_reviews]

plt.pie(sizes, labels=labels, autopct='%1.1f%%')
plt.title('Sentiment Distribution')
plt.show()

# ワードクラウド
wordcloud = WordCloud(stopwords='english').generate(' '.join(cleaned_reviews))
plt.imshow(wordcloud, interpolation='bilinear')
plt.axis('off')
plt.show()