The Effects of Interface Design of Recommendation Agents on Changing Negative eWOM

• Main Authors: Chia-Cheng Hsu, 許家誠
• Other Authors: Her-Sen Doong
• Format: Others
• Language: zh-TW
• Published: 2009
• Online Access: http://ndltd.ncl.edu.tw/handle/32694991078703897771

碩士 === 國立嘉義大學 === 資訊管理學系碩士班 === 97 === Since the Internet offers a large quantity of information, if costumers could not effectively choose or absorb the information, it would lead to “information overload.” So as to solve these problems, online shops have offered many mechanisms to help consumers. The most common mechanisms are eWOM (electronic Word-of-Mouth) and RA (Recommender Agent).However, to online shops which offer the two recommender systems mentioned above, it may probably lead to the problems of information inconsistency or information conflict. In such cases, the majority of online customers will tend to trust the eWOM. So this study aimed to effectively improve customers’ trust in RAs by means of the methods like avatar, explanation, virtual product experience, and find out how could recommender systems (eWOM vs. RAs) affect the negative or medium evaluations generated from eWOM under the circumstance of information inconsistency by using methods of interface design. This study used laboratory experiment and recruited 261 questionnaire.The result show that website presentation format will significantly enchance RAs perceived diagnosticity and trust, and finally will effect the intention of use RAs and purchase products.