Affiliation:
1. China Agricultural University
Abstract
The means of active recommendation is now increasingly applied to many different domains as it might meet users’ growing demands of personalized information service in process of selective purchasing for products, and research of recommendation technology based on user’s characteristics has been a key role in study area of active recommendation. However, investigations were performed to view there are still some deficiencies on existing methods. Therefore, an active recommendation method based on users’ static and dynamic characteristics is present and this paper proceeds as following. Firstly, feature model of user interests is constructed by analyzing static and dynamic data of target users for specific recommend. After that, neighbor users who have similar attributes with object user are found by reference to user model. Last, preference of intended user for other resources is reasonably predetermined by calculating interestingness with combination of neighbors, finally in this way can the main theory basis of recommendation comes into being. So in this paper a new approach based on users’ characteristics is provided, and it is shown with this method can effectively solve problem of shortage of satisfaction of traditional recommendation.
Publisher
Trans Tech Publications, Ltd.
Reference11 articles.
1. Abbattista F. Improving the Usability of an E-commerce Web Site through Personalization[A]. Proceedings of the Workshop on Recommendation and Personalization in Ecommerce, (2002).
2. Wu LH, Liu L. An Information Recommendation Method Based on Neural Networks[J]. Computer Engineering and Applications, 2005, 25.
3. Wu LH, Liu L, et al. Modeling User Multiple Interests by an Improved GCS Approach [J]. Expert Systems With Applications, 2005, 29: 757-767.
4. Adomavicius G. Alexander, and Tuzhilin. Toward the next generation of recommender systems: a survey of the state—of-the-art and possible extensions[J]. In IEEE Transactions on Knowledge and Data Engineering, (2005).
5. Choonho Kim and Jnntae Kim. A Recommendation Algorithm Using Multi—Level Association Rules[C], Proceedings of the IEEE/WlC International Conference on Web Intelligence(WT03), (2003).