A Proposed Method to Recognize the Research Trends using Web-based Search Engines

Abstract

This paper presents a novel approach to recognize research trends in a particular domain of research (i.e. Agent development) that is based on the number of data extracted from search engines. Several well-known mathematical and statistical theories have been used, from which a mathematical model has been derived to predict the agent development. The proposed solution attempts to convert the raw resulted number of documents found on the Internet (in PDF format) into useful information, and fit a curve representing the number of searched data of every year. A prototype has been designed, to search multi-keywords at the same time and collect the required data automatically to eliminate the useless data, before converting it into usable data.