基于文档词典的文本关联关键词推荐技术

Associated Text Entry Recommendation Technology Based on the Document Dictionary

针对搜索引擎只能检索出用户给定关键词的相关记录,无法搜集用户潜在感兴趣的关键词的问题,提出一种基于文档词典的新闻关联关键词推荐技术。首先采用词频向量模型算法表示语料库中文档,然后对给定输入的关键词使选择与之关联度最高的文档;最后使用TextRank算法选取筛选出文档中值最高的N个关键词,这些关键词即为关联关键词。理论和实验表明,该算法能够有效地根据关键词推荐与之相关的关联关键词。

In view of the search engine can only retrieve the relevant records of the user＇s given keywords, cannot collect the user potential interest keyword question, and proposes one kind of news relevance keyword recommendation technology based on the document dictionary. Firstly, uses the word frequency vector model algorithm to represent the document, then the keyword of the given input is chosen to select the most relevant documents. Finally, uses the Text-Rank algorithm to select the highest value of the N keywords in the document, these keywords are associated keywords. The theoretical and experimental results show that the proposed algorithm can be used to recommend the related key words effectively.