基于电子商务同类商品的推荐算法研究

Research on Similar Products Recommendation Algorithm Based on Electronic Commerce

个性化推荐算法是电子商务系统的研究热点。文中给出一种基于同类商品的推荐算法,使用户在购买商品时,快速得到性价比高的同类产品,提高系统的服务能力。算法针对同类产品,将供应商名称、商品价格、购买人数、收货人数、用户评论5个参数作为推荐指标,在充分论证的基础上,确定了各个指标的权重,在此基础上进行了数据建模。为验证该模型的正确性,抓取了2014年2月1日淘宝网(SAMSUNG/三星s7898)的产品列表,根据销售情况,选取前67个商家的销售情况进行实验。结果表明该模型客观、有效。目前,关于个性化的推荐算法较多,但针对于同类商品的推荐算法的研究成果相对较少,该推荐算法的实施可减少用户查找满意商品的难度,提高系统的服务水平。

Personalized recommendation algorithm is a hot issue in the study of the electronic commerce system. A recommendation algorithm based on similar products is presented in this paper,by which users in the purchase of goods can quick get cost- effective products and improve the service ability of the system. This algorithm aims at similar products,the five arguments including supplier name,commodity prices,the number of purchase,receiving the number,user reviews are selected as recommended indexes. On the basis of sufficient demonstration,the weight of each index is determined and data model is established. To test and verify the correctness of the model,a experiment is conducted according to the sales of the first 67 of the dealer of the list in Taobao product（ SAMSUNG / SAMSUNG s7898）on February 1,2014. The results showthat the model is objective and effective. At present,there are more personalized recommendation algorithms,but the research achievements of recommendation algorithm proposed for the similar goods are relatively small,the implementation of the recommendation algorithm can reduce the difficulty of users finding satisfactory goods,and improve the service level of the system.