Research On Tag Recommendation Based on Multiple Keywords

摘要

The user inquiry system is a tool that we often use in our daily life. The effect of a inquiry system directly affects the user's experience. But the user will enter multiple keywords in the query to express what they really want to search, but this is usually a lot of noise caused by the impact of individual keywords, so this paper proposes a keyword query intention of mixed probability model, the model can be recommended result of query words together with the query intention fusion, and thus to minimize noise on the result of recommended keywords. This article also add tags to key words, as the query keywords supplement,. This method can not only make up for the noise interference of multiple keywords, but also calculate the similarity between the tag and the tag that has been marked on the web page, so as to judge the relevance between the user's query intention and the web page content more accurately. Finally, the results that are more in line with the user's query interest are ranked in the front position. Experimental results show that the proposed method is more effective.