基于不完全信息查询发现推荐规则的一种方法

摘要

In case-based reasoning （CBR） systems for product recommendation, the retrieval of acceptable products based on limited information （incomplete information） is an important and challenging problem. This paper investigateds an approach to the discovery of recommendation rules to support a rule- based approach to the retrieval of recommended cases. While having the potential to provide greater coverage than decision-tree and nearest neighbor approaches and enabling recommendations to be easily justified, the most exact rules discovered by our discovery algorithm offer obvious improvement in terms of their reliability when applied to incomplete queries.%在CBR推荐系统中，基于受限的信息（不完全信息）进行检索，得到可接受的产品是一个重要的、具有挑战性的问题。文章提出了一种发现推荐规则的方法：最优规则推荐算法用来检索事例库。该方法应用于不完全查询中，在发现的规则数量上与NN方法发现的规则基本相同，但这些规则有较高的事例覆盖率。跟决策树方法相比，该方法发现的规则数量少，更容易解释，而且有较高的事例覆盖率。