The Effect of Conformity-based Filtering on Information Load

• Main Authors: Karen, 林若萍
• Other Authors: Wang, Jyun-Cheng
• Format: Others
• Language: zh-TW
• Published: 2002
• Online Access: http://ndltd.ncl.edu.tw/handle/07270935041139314810

碩士 === 國立中正大學 === 資訊管理學系 === 90 === This study attempts to extend the meaning of personalization and argues that not only personal information needs but also emotional or mental needs aroused by outside influences need to be taken into account in such a redefinition. This research assigns a new characteristic that can be included into the filtering out unnecessary information: the conformity behavior. Conformity means that people will tend to converge on similar behavior because they are affected by social norms. This study compares the effects of four personalization mechanisms on subjective decision quality. The results show that pure conformity is better than target conformity. Target conformity is no significant different from collaborative filtering. And collaborative filtering is better than no filtering. The implications for the results are very positive and important. As the operational meaning for conformity is simply showing the page retrieval count for each piece of information, it is inexpensive to implement. More importantly, in doing conformity-based filtering, web site owners can elude the accusation of jeopardizing privacy, by which personalization systems have been under attack.