Scholarly record

COMPRESSION OF ANSWERS (SERP) FROM SEARCH ENGINE: ALTERNATIVE RESEARCH WITH MULTILINGUAL REQUEST REFORMULATION

Omar Larouk, M. Hassoun

Abstract

The Internet has become a very important source of information for users (students, researchers, general public, etc.) and commercial enterprises. This propulsion of data is related to the diversity of information demands and the need for versatile knowledge. Access to digital data via web services is facilitated by search tools. Following a user request, the search engine has a long list of bibliographic references and web pages. The efficient selection of relevant documents is very difficult given the low accuracy of the list. Typically, the user visits the first page referenced in the list, but only consults the following pages. In this article, we describe the problem of the query language when we search for information, via an automatic system (such as search engines) and the problems related to the compression of indexes and the economy of responses (SERPs) provided by the index of each tool experienced. We propose to compare the answers SERP (Search Engine Result Page) of four tools (Google, Bing, Yahoo and Yandex) and to test formal and natural language queries. We will calculate the relevance of the answers given and the number of relevant answers according to the predefined qualitative criteria on the presence of the descriptors in the URL, and in the metadata. This alternative search, on queries comprising several words, shows that our combinatorial methods based on the coupling of questions reconstruction with logical operators (BOOLE) and the proximity operators allow a filtering of the responses (SERP), and therefore a reduction of documentary noise related to inadequate information. We will give a ranking of the relevance of the indexes of the search engines analyzed.

Publication Impact Profile

No metrics available.

PLUMX see details

DOI resolver: doi.org/10.5593/sgemsocial2018/3.6/S14.017

Publication details

Back to publication list