Efficient web object caching through probabilistic modeling

Abstract

Web object search engines are a new breed of Web IR systems, wherein, the focus is to provide relevant results, such that, the result set has limited influx of unnecessary information. To achieve this said goal, the document information is extracted, and reorganized in the Web object repository, wherein, Web objects are logical entities such as-cities, authors, stadiums etc. Caching has become an integral part of modern Web search engines. Many efficient caching schemes have been proposed for this domain. However, the same attention is lacking for Web object search engines. The contemporary solution in the literature is quite primitive, and lacks the required effectiveness. In this work, a new caching mechanism based on probabilistic model is presented. The proposed caching mechanism is evaluated empirically along with the contemporary technique. Empirical results exhibit superior performance of the proposed scheme in-terms of cache hit ratio and query execution time.