BrandCultivation/models/item2vec/item2vec.py

import joblib
from database.dao.mysql_dao import MySqlDao
from models.item2vec import Item2VecDataProcess
import numpy as np
from gensim.models import Word2Vec

from models.rank.data.config import ProductConfig
from models.rank.data.utils import sample_data_clear
import pandas as pd
from sklearn.metrics.pairwise import cosine_similarity
from tqdm import tqdm

class Item2Vec:
    def __init__(self, city_uuid):
        self._load_data(city_uuid)
        self._load_model()
    
    def _load_data(self, city_uuid):
        """加载特征sentence"""
        data_processor = Item2VecDataProcess(city_uuid)
        self._tokens_map = data_processor.generate_tokens()
        self._tokens = [item["token"] for item in self._tokens_map]
            
    def _load_model(self):
        self._model = Word2Vec(
            self._tokens,
            vector_size=64,
            window=3,
            min_count=1,
            sg=1, # skip-gram
            workers=1,   # 固定为1，保证多线程不引入随机性
            seed=123456,     # 固定随机种子，确保结果可复现
            epochs=20,
            sample=0.0000001
        )
        
    def token_to_vector(self, tokens):
        """将token转换为vector"""
        vector = [self._model.wv[token] for token in tokens if token in self._model.wv]
        return np.mean(vector, axis=0) if vector else np.zeros(self._model.vector_size)
    
    def item_to_token(self, item):
        token = []
        for col in ProductConfig.FEATURE_COLUMNS:
            if col == 'product_code':
                continue
            else:
                token.append(f"{item[col].strip()}")
        
        return token
    
    def get_similarity_map(self, product):
        """获取目标卷烟与所有卷烟的相似度"""
        product = product.squeeze().to_dict()
        token = self.item_to_token(product)
        vector = self.token_to_vector(token).reshape(1, -1)
        
        similarity_map = []
        for item in tqdm(self._tokens_map, desc="正在计算卷烟相似度..."):
            target_product_code = item["product_code"]
            torget_token = item["token"]
            target_vectot = self.token_to_vector(torget_token).reshape(1, -1)
            
            similarity = cosine_similarity(vector, target_vectot)[0][0]
            
            similarity_map.append(
                {
                    "target_product_code": product['product_code'], 
                    "product_code": target_product_code,
                    "similarity": similarity
                }
            )
        similarity_map.sort(key=lambda x: x["similarity"], reverse=True)
        
        return similarity_map
        
            

        
        
if __name__ == "__main__":
    dao = MySqlDao()
    city_uuid = "00000000000000000000000011445301"
    product_id = "350139"
    
    product = dao.get_product_by_id(city_uuid, product_id)[ProductConfig.FEATURE_COLUMNS]
    product = sample_data_clear(product, ProductConfig)
    model = Item2Vec(city_uuid)
    sims = model.get_similarity_map(product)
    sims = pd.DataFrame(sims)
    product_info = dao.load_product_data(city_uuid)[ProductConfig.FEATURE_COLUMNS]
    sims = sims.merge(product_info, on="product_code", how="inner")
    sims.to_csv("./data/product_similarity.csv", index=False)