锦中融合门户系统

# 导入必要的库

import pandas as pd

from sklearn.neighbors import NearestNeighbors

# 模拟用户-物品评分数据

data = {

'user_id': [1, 1, 2, 2, 3, 3],

'item_id': [101, 102, 101, 103, 102, 103],

'rating': [5, 4, 3, 2, 4, 5]

}

df = pd.DataFrame(data)

# 构建用户-物品矩阵

user_item_matrix = df.pivot_table(index='user_id', columns='item_id', values='rating').fillna(0)

# 使用KNN进行相似度计算

model = NearestNeighbors(n_neighbors=2, algorithm='brute', metric='cosine')

model.fit(user_item_matrix)

distances, indices = model.kneighbors(user_item_matrix.values)

# 推荐相似用户喜欢的物品

def recommend_items(user_id):

user_index = user_item_matrix.index.tolist().index(user_id)

similar_users = indices[user_index][1:] # 排除自己

recommended_items = set()

for idx in similar_users:

similar_user_id = user_item_matrix.index[idx]

items_rated_by_similar_user = user_item_matrix.iloc[idx].to_dict()

for item, rating in items_rated_by_similar_user.items():

if rating > 3 and item not in user_item_matrix.columns:

recommended_items.add(item)

return list(recommended_items)

# 示例：为用户1推荐物品

print("为用户1推荐的物品:", recommend_items(1))

# 使用Surprise库实现SVD推荐系统

from surprise import Dataset, Reader, SVD

from surprise.model_selection import train_test_split

# 加载数据

data = Dataset.load_builtin('ml-100k')

# 划分训练集和测试集

trainset, testset = train_test_split(data, test_size=0.25)

# 使用SVD算法训练模型

model = SVD()

model.fit(trainset)

# 预测用户对物品的评分

predictions = model.test(testset)

# 输出预测结果

for p in predictions[:5]:

print(f"用户 {p.uid} 对物品 {p.iid} 的预测评分为: {p.est}")

# app.py

from flask import Flask, request, jsonify

import pandas as pd

from sklearn.neighbors import NearestNeighbors

app = Flask(__name__)

# 模拟用户-物品评分数据

data = {

'user_id': [1, 1, 2, 2, 3, 3],

'item_id': [101, 102, 101, 103, 102, 103],

'rating': [5, 4, 3, 2, 4, 5]

}

df = pd.DataFrame(data)

user_item_matrix = df.pivot_table(index='user_id', columns='item_id', values='rating').fillna(0)

model = NearestNeighbors(n_neighbors=2, algorithm='brute', metric='cosine')

model.fit(user_item_matrix)

def recommend_items(user_id):

user_index = user_item_matrix.index.tolist().index(user_id)

similar_users = model.kneighbors(user_item_matrix.values[user_index], n_neighbors=2)[1][0][1:]

recommended_items = set()

for idx in similar_users:

similar_user_id = user_item_matrix.index[idx]

items_rated_by_similar_user = user_item_matrix.iloc[idx].to_dict()

for item, rating in items_rated_by_similar_user.items():

if rating > 3:

recommended_items.add(item)

return list(recommended_items)

@app.route('/recommend', methods=['POST'])

def recommend():

user_id = request.json.get('user_id')

recommendations = recommend_items(user_id)

return jsonify({'recommendations': recommendations})

if __name__ == '__main__':

app.run(debug=True)

# search_api.py

from flask import Flask, request, jsonify

from elasticsearch import Elasticsearch

app = Flask(__name__)

es = Elasticsearch(hosts=['http://localhost:9200'])

@app.route('/search', methods=['GET'])

def search():

query = request.args.get('q')

result = es.search(index="content", body={"query": {"multi_match": {"query": query, "fields": ["title", "content"]}}})

hits = result['hits']['hits']

return jsonify([{'title': hit['_source']['title'], 'content': hit['_source']['content']} for hit in hits])

if __name__ == '__main__':

app.run(debug=True)