融合类别与评分的茶产品协同过滤推荐研究

Study on Collaborative Filtering Recommendation of Tea with Fusion Category and Score

摘要: 个性化推荐作为当前电子商务领域研究的热点之一，也是未来茶产品电子商务发展的必然趋势。论文针对现有个性化推荐中将茶叶作为普通商品推荐，没有考虑其自身农产品特性，同时对传统协同过滤推荐技术仅考虑消费者评分导致的推荐精度不高等问题，提出将评分与商品类别相融合的方式实现对茶叶的协同过滤推荐。通过真实数据集上的实验表明，新的推荐方法提高了茶产品个性化推荐的准确度，对实现茶叶电子商务的个性化营销具有一定的意义。

Abstract: As one of the main content of the current research and application in the electronic commerce field, personalized recommendation will be the inevitable development in the future of tea e-com- merce applications. The existing personalized recommendation technology regarded tea as common goods recommending to consumers, but ignored the agricultural characteristics. And aimed at the problems such as the few recommendations led by the traditional personalized recommendation only considering consumer ratings, this paper proposed a method of the construction of tea consumer network fusing tea category similarity. The experiment on the basis of real data sets indicated that, the new recommendation method improves the personal recommendation accuracy for the tea products, having a certain significance to realize personal marketing of tea e-com- merce.