基于日志挖掘的检索推荐系统

A Recommendation System for Information Retrieval Based on Web Logs Mining

目的为了有效地预测用户在信息检索过程中可能点击的检索结果,从而进行网页的智能推荐.方法采取网络日志挖掘的技术,通过词频信息和知网(HowNet)中词的概念计算模型计算网页文档间的主题相关度,再将该语义信息与统计模型计算的条件概率值相结合,以此作为网页推荐的依据.结果提出了一种检索推荐统计模型,并构建了相应的原型系统,实验表明该方法显著提高了推荐系统的准确率.结论这项技术有效地提高了推荐结果与用户信息需求的相关程度,使推荐系统的性能获得了较大地提高,可以很好的应用于信息检索的智能推荐服务领域.

The Web page recommendation systems can effectively predict users＇ next clicking results in information retrieval process and the research will be beneficial to many applications ranging from intelligent recommendation to improving effectiveness of search engines. In this paper, in order to deal with the problem of lack of semantic processing in present systems, the technology of Web log mining has been adopted to use word frequency and the concept relevancy model of HOWNET to compute document relevancy, and the result is used to guide the process of pages recommendation. In the end, a relevancy-based recommendation system based on query logs mining is proposed, which combines document relevancy calculation with the method of statistical language model. Furthermore, the prototype system has been built and the experiment showed that this method significantly improved the accuracy of recommendation systems. In conclusion, this method outperforms other models in the web page recommendation systems and overcomes the problem of the lack of effective semantic process. The performance of recommendation system has been improved greatly and the technique can be widely used on the intelligent recommended services in information retrieval field.