H.O. Nyongesa and B. Gopolang (Botswana)
user information needs modelling, interactive evolutionary learning, information relevance, adaptive information retrieval
The amount of information and variety of its sources is becoming an obstacle to obtaining information that accurately matches user information needs. Two key factors affect the accuracy of matching user information needs against the retrieved information (retrieval effectiveness). First, users often do not represent their information needs in the form that, in fact, represents their actual information needs. Second, the relation between a document's subject matter and a user’s information need (relevance) is highly subjective between different users. This paper presents an approach for improving retrieval effectiveness from document databases. It is proposed that the retrieval effectiveness can be improved through learning by applying computational intelligence techniques combining qualitative user relevance judgment with quantitative metrics of the relevance of retrieved documents.
Important Links:
Go Back
