摘要
提出一种基于认知复杂度度量的文本推荐模型。已有的认知复杂度评价方法主要用于评价单一文本的认知复杂度,对此方法进行拓展,将它用于文本集的认知复杂度评价。在推荐模型中,通过对用户查看文章序列的分析,然后使用文本集认知复杂度评价方法查找文章进行推荐,使得用户获得的推荐文本集合更符合认知的规律,更易于理解。实验结果表明文本集认知复杂度评价方法的合理性,并通过比较说明了使用这种推荐模型将使用户获得更加易于理解的文本推荐集合。
A document recommendation model based on cognitive difficulty measurement is presented. The existing cognitive dif- ficulty measurement algorithm is designed to work on single document. The algorithm is improved to measure the cognitive diffi- culty of document set. In the recommendation model of this paper, according to the documents which the user has read, a recom- mended document set is given through cognitive difficulty measurement. The given documents will help the reader to understand the document more easily. At the end, the reasonability of cognitive difficulty measurement algorithm of document set is proved by experiment. Through comparison, the recommendation model is proved to be more easily to cognitive for the readers.
出处
《计算机工程与设计》
CSCD
北大核心
2012年第10期3990-3994,共5页
Computer Engineering and Design
关键词
推荐模型
认知复杂度
主题提取
概念计算
文本聚类
recommendation model
cognitive difficultyl topic extract
concept computing
document cluster
作者简介
许霄峰（1983-），男，上海人，硕士研究生，研究方向为文本挖掘和数据挖掘；
徐炜民（1949-），男，上海人，教授，博士生导师，研究方向为计算机系统结构和人工智能等。E-mail：haha21333@gmail．com
