Intelligent mobile agent framework for searching popular e-advertisements

Abstract

e-Advertising is the rapidly growing e-commerce application which acts a significant entity for user browsing behavior/patterns and is employed by majority of ad platforms to evaluate the ad selection process for choosing the right product. Existing search engines adopt other link structure/content-oriented and do not consider the browsing patterns. In link structure, rank scores are applied evenly irrespective of the links as target web pages are self-descriptive and use links for navigating. In a content-oriented, the web page lacks in having the rich content description to match with the search query. Thus, generating the top priority list for the user query is a difficult task. Here, mobile agent architecture is proposed to perform the search engine process by tracking the ad relevance and to estimate the probability of views/clicks to distribute the rank scores based on the ad popularity. Mobile agents extend to apply classification technique to classify the ad list into three classes and display the most relevant ads which improve the result list.