X. Fan, F. Sheng, G. Thomas, and P.A. Ng (USA)
Document retrieval, knowledge representation, machine learning
The effectiveness and efficiency have been the major challenges in the document and information retrieval field. Personal information systems can be built based on user's special preference and interests on documents, which can be used to dramatically improve the efficiency and accuracy of document retrieval. This paper presents a flexible dual-model framework that gets user involved in document modeling. Being familiar with the document model, user can specify complex and precise queries. A knowledge-based approach to document filing and retrieval is presented. To support such a flexible document model, a knowledge representation model is proposed to store the domain knowledge. Machine learning algorithms are proposed to acquire the domain knowledge.
Important Links:
Go Back
