A re-examination of relevance: toward a dynamic, situational definition
Information Processing and Management: an International Journal
Combining the evidence of multiple query representations for information retrieval
TREC-2 Proceedings of the second conference on Text retrieval conference
Combining multiple evidence from different properties of weighting schemes
SIGIR '95 Proceedings of the 18th annual international ACM SIGIR conference on Research and development in information retrieval
Modern Information Retrieval
A survey on the use of relevance feedback for information access systems
The Knowledge Engineering Review
An outranking approach for information retrieval
Information Retrieval
Introduction to Information Retrieval
Introduction to Information Retrieval
Hi-index | 0.00 |
Retrieving relevant items as a response to a user query is the aim of each information retrieval system. But `without an understanding of what relevance means to users, it is difficult to imagine how a system can retrieve relevant information for users' [1]. In this paper, we try to capture what relevance is for a particular user and model his profile implicitly considering his non declared preferences that are inferred from a ranking of a reduced set of retrieved documents that he produces. We propose an ordinal regression based model for interactive ranking which uses both the information given by this subjective ranking, as well as the multicriteria evaluation of these ranked documents, to adjust optimally the parameters of a ranking model. This model consists of a set of additive value functions which are built so as they are as compatible as possible with the subjective ranking. The preference information used in our model requires reasonable cognitive effort from the user.