A hybrid approach for user profiling
- 1 January 2003
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 9 pp.-103.2
- https://doi.org/10.1109/hicss.2003.1174242
Abstract
With the growing amount of information being digitized, users find it difficult to obtain the most relevant information that is hidden in the deluge of information returned to them by search engines. In this paper, we describe a conceptual model using a hybrid approach based on user profiling approaches and information filtering techniques that can alleviate the problem of information overload. We discuss the concepts, namely static content profiling, static collaborative profiling, dynamic content profiling and dynamic collaborative profiling, followed by the design and implementation of a library search facility, which has employed these concepts. Preliminary experiments conducted have shown that users have benefited from our prototype system. It appears that there is value in employing these new concepts that we have proposed in the design and implementation of future information retrieval system for better retrieval.Keywords
This publication has 6 references indexed in Scilit:
- INVAID: an Intelligent NaVigational AID for the World Wide WebPublished by Institution of Engineering and Technology (IET) ,1999
- Context-sensitive filtering for browsing in hypertextPublished by Association for Computing Machinery (ACM) ,1998
- GroupLensCommunications of the ACM, 1997
- Recommender systems for evaluating computer messagesCommunications of the ACM, 1997
- User Models and Filtering Agents for Improved Internet Information RetrievalUser Modelling and User-Adapted Interaction, 1997
- BIRCHACM SIGMOD Record, 1996