Title :
Formulation of the term refinement problem for user-oriented information retrieval
Author :
Deogun, Jitender S. ; Raghavan, Vijay V. ; Rhee, PhillKyu
Author_Institution :
Dept. of Comput. Sci., Nebraska Univ., Lincoln, NE, USA
Abstract :
A framework called the user-oriented information retrieval system is introduced. The framework enables the retrieval system to be adaptive in the sense that not only the structuring and organization of documents, but also their representation can be dynamically modified through feedback. An important problem that arises in this context is one of determining the significance for retrieval of index terms assigned to documents. This problem, called the term refinement problem, is formulated rigorously. It is shown that there is a direct correspondence between this problem and the well known (0,1)-knapsack problem
Keywords :
information retrieval; pattern recognition; cluster based retrieval system; cluster characterization; document clustering; index terms; knapsack problem; response relevance feedback; significance; term refinement problem; user-oriented clustering; user-oriented information retrieval; Adaptive systems; Approximation algorithms; Clustering algorithms; Computer science; Feedback; Genetics; Heuristic algorithms; Indexing; Information retrieval;
Conference_Titel :
AI Systems in Government Conference, 1989.,Proceedings of the Annual
Conference_Location :
Washington, DC
Print_ISBN :
0-8186-1934-1
DOI :
10.1109/AISIG.1989.47307