作者: F.B. Pereira , E. Costa
关键词:
摘要: Queries that are not indicative of real information needs a major problem for retrieval systems. In this work we study how individual learning helps adaptive agents, when searching in distributed environment, to modify incomplete queries order improve their retrieving performance. Two procedures, occurring two different levels, proposed and effect is studied several situations. Preliminary results show changes induced by the query vector provide an important advantage enable them make correct decisions about deal with problem.