A Study on Collaborative Information Filtering Systems Using the Fuzzy Linguistic Analysis

• Main Authors: Yao-te Tsai, 蔡耀德
• Other Authors: Lai-hsi Lee
• Format: Others
• Language: zh-TW
• Published: 2006
• Online Access: http://ndltd.ncl.edu.tw/handle/41704657468742363627

碩士 === 國立屏東商業技術學院 === 資訊管理系 === 94 === This study has two purpose: one is to search the nearest neighbor in collaborative filtering system using the fuzzy linguistic analysis. In the beginning of the filtering system a questionnaire of information involvement is set to collect data of new users. Then the fuzzy linguistic analysis is used to analysis the difference of apprehension among users in the questions of the questionnaire. Additionally, a difference index between two users is then calculated as the comparison base for searching the nearest neighbor. An example is used further to illustrated the procedure of the method and shows possibility for searching the nearest neighbor. The second is to filter information to that user from the nearest neighbor of that user in collaborative filtering system. The purpose of information filter process is provide valuable information that user’s need. In which users are asked to provide necessary explicit evaluation information to the system using the fuzzy linguistic analysis. So that the collaborative filtering system based on the users’ evaluation for the topic to propose recomdation and conform to preference of users. In this thesis, I try to apply collaborative filtering system in web sites to help users finding their favorite information topic. In summary, the contributions of this thesis include creating a new way of information filter and achieving the goal of raising predictive accuracy.