An Interactive Agent System for supporting Knowledge-based Recommendation － A Case Study on an e-Novel Recommender System

• Main Authors: Ruei-tang Huang, 黃銳堂
• Other Authors: Jim-Min Lin
• Format: Others
• Language: zh-TW
• Published: 2009
• Online Access: http://ndltd.ncl.edu.tw/handle/32378767459357469622

碩士 === 逢甲大學 === 資訊工程所 === 97 === With the fast development of e-commerce, the Internet is full of a large number of complex information regarding diverse products. Many researches tend to design a recommender system to assist users by identifying interesting products and services in the situations where the number and complexity of offers beyond the user’s capability to go through all the information for a decision. The knowledge-based recommendation is a scheme emphasizing the needs that through user interaction to guide the users to contribute their own needs and create a user preference model. In the past, agents commonly play the role of information retriever in the recommender systems. Agents have the capability of autonomy and cooperation for assisting the collection of user information and analysis of user''s preferences. However, rare researches used the embodied conversational agents to interact with users. Therefore, we design an interactive agent system for supporting knowledge-based recommendation in this study. In addition to recommend product through the knowledge from the interaction with the user, the knowledge-based approach emphasizes the system should also explain the recommended results to the user. For this reason this study used embodied conversational agents to do the promotion and explanations of the recommendations. As a result, our proposed system could reach the purpose of advertising products and receiving user feedback. In this paper, we implement the system in an e-novel web site, called Angel City, in order to recommend e-novels and demonstrate the usability of our system.