摘要
在网页浏览推荐任务中,如何利用网页内容选取合适的推荐关键词是具有挑战性的研究热点.为了实现有效的关键词推荐方法,利用大规模的真实网络用户浏览行为数据,以及相关提取算法和新词发现算法实现并比较了基于领域关键词提取技术和基于查询词候选集合的关键词推荐方法.实验结果证明,2种方法都能够有效地表征用户信息需求,而第1种推荐方法的准确率更高,具有更好的推荐性能.
It is very challenging when conducting research and it is especially difficult as it pertains to website brow sing and recommendation system task for selection of suitable keyword usage. This research study will focus on proper use of website browsing and recommendations on how to select keywords for conducting research. The chal lenge is to leverage user behavior features, as well as develop an effective keyword＇s recommendation content page. The implementation of a comprehensive user browsing data, relevant extraction algorithm and algorithm finding methods for new keywords were examined in the research study. The research study also proposed additional, key word recommendation methods utilizing large-scale and related algorithm approaches for domain-specific keyword extraction technology and a query keyword candidate set were compared. The experiment results confirm both meth ods demonstrate that they satisfy users＇ information demand. However, the keyword recommendation methods show a significant performance improvement in effectiveness. The keyword recommendation method has a higher accuracy and better recommendation performance.
出处
《智能系统学报》
北大核心
2012年第5期398-403,共6页
CAAI Transactions on Intelligent Systems
基金
国家自然科学基金资助项目(60736044,60903107,61073071)
高等学校博士学科点专项科研基金资助项目(20090002120005)
关键词
浏览推荐
关键词推荐
关键词提取
网页关键词
browsing recommendation
keyword recommendation
keyword extraction
web keywords
