实验四：基于内容的推荐

代码

import pandas as pd
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.metrics.pairwise import cosine_similarity
from sklearn.metrics.pairwise import pairwise_distances
import numpy as np

news_df=pd.read_csv('C:/Users/Administrator/Desktop/uci-news-aggregator.csv')

vectorizer=TfidfVectorizer(stop_words='english')
news_vectors=vectorizer.fit_transform(news_df['TITLE'])

def cosine_recommendation(title):
    title_vector=vectorizer.transform([title])
    cosine_similarities=cosine_similarity(title_vector,news_vectors).flatten()
    related_docs_indices=cosine_similarities.argsort()[::-1]
    return list(news_df['TITLE'].iloc[related_docs_indices][:10])

#测试案例
test_title='Microsoft releases Windows 10 update'
print('基于 cosine 相似度的推荐结果：',cosine_recommendation(test_title))

结果

数据集

见上述资料