| 基于商品属性与用户聚类的个性化服装推荐研究 |
| |
| 引用本文： | 艾黎.基于商品属性与用户聚类的个性化服装推荐研究[J].现代情报,2015,35(9):165-170. |
| |
| 作者姓名： | 艾黎 |
| |
| 作者单位： | 武汉大学信息管理学院, 湖北 武汉 430072 |
| |
| 摘 要： | 淘宝网作为电子商务时代最大的网上零售平台,为用户提供越来越多的商品与服务的同时,也出现了信息过载等一系列问题。鉴于此,本文提出了基于商品属性与用户聚类的个性化服装推荐方法,通过用户个人信息与对商品的评价,计算用户之间的相似度,进行聚类分析。与此同时,将商品化整为零,通过商品属性来计算商品的相似度,得到top-N相似列表。以此,综合商品与用户两者的权重值,实现为用户提供个性化的商品推荐,解决用户面对信息过载的难题,为用户节省精力,提高用户的购物体验。针对某一淘宝网店铺,本文提出了适合的混合推荐算法,并通过搜集实际数据进行了实证研究,对推荐结果进行准确性评价。
|
The Research on Personalized Recommendation Based on Commodity Attribute and User Clustering |
| |
| Authors: | Ai Li |
| |
| Institution: | School of Information Management, Wuhan University, Wuhan 430072, China |
| |
| Abstract: | As the largest online retail platform in the era of e-commerce, Taobao provides users with more and more goods and services, but it also has a series of problems such as information overload.In this view, the paper proposed a personalized clothing recommendation method based on commodities' attributes and users' clustering.According to user's personal information and his or her comment of the commodity, the paper could calculate the similarities between users, then divide them into different clusters.Meanwhile, the paper described the commodity as a set of attributes and calculate similarities of the products.Then the paper got a list of top-N similar products.With the weights of commodity's similarities and user comments, it provided users with personalized commodity recommendations, solving the problem of information overload.It's aimed to save energy, improve the user's shopping experience.Take the example of one Taobao shop, empirical research is carried out by collecting the actual data to evaluate the precision of hybrid recommendation algorithm.And the results were not bad. |
| |
| Keywords: | |
|
| 点击此处可从《现代情报》浏览原始摘要信息 |
| 点击此处可从《现代情报》下载免费的PDF全文 |
|
