AN EFFICIENT INFORMATION RETRIEVEAL FROM PLURAL INDEPENDENT DATABASES PARTIALLY UNRELIABLE

摘要

In this paper, we propose an efficient information retrieval scheme that collects information from a number of sources scattered over a large-scale network while avoiding error interfusion. It was designed to obtain each portion of data that is newly acquired by more reliable sources and to configure all accurate and necessary information. Furthermore, in order to illuminate our scheme's efficiency, we conducted computer simulations and compared its performance to that of other conventional schemes. The results indicate that the proposed scheme, compared to the other schemes, achieved a higher accordance rate between the actual information and configured information. In addition, the simulation results also proved that the proposed scheme is adequate for the conservation of data freshness.