A STUDY ON THE SIMILARITY IN THE RESULTS OFFERED BY THE MAIN SEARCH ENGINES ON THE WEB

Abstract

The huge increment of the available information on the web added to a progressive augment of its use as main source of data has caused that the utilization of information retrieval systems -IRS- in the web (also known as search engines), takes on a higher relevance. Even though these tools carry out an essential work, no less necessary is the task of evaluating them and studying the data that they provide us. This research proposes to analyze the main search engines of the market through a direct comparison of the URLs returned from a heterogeneous set of queries to intend to find out the current degree of similarity of the responses. For this purpose we have developed a metasearch engine which will let us make our search experiment and calculate the distances in a quick and interactive way. Finally, we will try to get valid conclusions of the results contrasting them with data from other researches and from our own experiment, what will demonstrate the high degree of divergence in the responses of the IRS.