Indian Journal of Science and Technology

Abstract

Background/Objectives: Provide better search results and improve overall search experience of the users while searching information over the Internet. Most of the commonly used search engines today utilize link structure based page ranking for ordering of web pages and give no consideration to web page’s relevance and user-feedback etc. and suffer from topicdrift problem. As a result users seldom find accurate, precise and desired information on first page of the result-set. The users have to scan through the large result-set refining it manually which adversely affects the overall search experience of the user. Therefore there is a need to provide better ranking solutions. Methods/ Statistical Analysis: The methodology used for ranking a web page in the proposed algorithm includes ranking based on web page characteristics, user explicit feedback and preferences. The algorithm computes page rank of a web page based on number of times of occurrence of user specified Keyword on the web page, web page’s priority, its rating and popularity as specified by user feedback The algorithms also attempts to match user preferences in terms of domain and type of information of the web page user are interested in for page ranking. Findings: The proposed algorithm promises to provide best web pages in conjunction with user needs and interest, thereby improving the overall quality of the search results and user satisfaction. Application: The proposed algorithm can be used in Search Engines to provide better user specific results.

Keywords: Information Retrieval, Page Ranking, Search Engine, User-feedback, User-Preferences, User-satisfaction