基于粗糙集理论的金融信息推荐算法研究.docx

计算机研究与发展 计算机研究与发展 ISSN 1000—1239／CN 11．1777／TP Journal of Computer Research and Development 44(Suppl．)：262～270，2007 基于粗糙集理论的金融信息推荐算法研究 辛治运1’2 顾 明2 马兆丰3 1(清华大学计算机科学与技术系北京 100084) 2(清华大学软件学院北京100084) 3(北京邮电大学信息工程学院 北京 100876) (mzf@tsinghua．edu．ca) Recommendation Algorithm Based on Rough Set for Specific Domain Information Service Xin Zhiyunl．一，Gu Min92，and Ma Zhaofen93 1(Department of Computer Science and Technology，Tsinghua University，Beijing 100084) 2(School of Software，Tsinghua University。Beijing 100084) 3(School of Information Engineering，BeOing University ofPosts and Telecommunications，Beijing 100876) Abstract Information recommendation is the information service technology which recommends relative for according users’interest．Present information recommendation technology mainly provides information service in global information domain．However，the needs special domain information．For finding recommendation method which satisfies special domain information need，a recommendation algorithm which combines rough with content—filtering technology for special domain information service is presented in this paper．Rough theory is advantageous dealing with uncertainty and vagueness information．First，documents classification based rough theory is done，which includes some steps such data preconditioning，features selection，redundant attribute reduction，rules production and finally classification result．Then special domain information recommendation based content—filtering is realized．From the experiment result，the efficiency of algorithm is improved．Compared with the common classification algorithm，the special domain information recommendation algorithm based content-filtering makes the precision ratio and recall ratio higher．It recommend some important information precisely for certain special Key words rough set；document classification；content—filtering；information recommendation 摘要信息推荐是根据用户需