Developing a More Effective Ranking Algorithm with Query Analysis than Google's PageRank

Overview

Objectives/Goals The Researcher developed a new ranking algorithm of Web pages. This algorithm ranks Web pages by counting the number of query keywords and analyzing the distribution of the query keywords within a page. The objective was to see if the new algorithm is more effective than Google's PageRank in determining the relevance of a page. Methods/Materials Wrote a VB program that retrieves Google's top 500 -1000 Web pages with a query of 3 - 4 keywords and ranks them using the new algorithm. Randomly mixed the top 10 pages ranked by the researcher's algorithm with Google's top 10 pages and let 11 people evaluate these 20 web pages regarding their relevance to the query. Also, repeated above steps using other 10 queries in different subjects and let three people determine their relevance to the queries. Results The users' average evaluation scores showed that the researcher's algorithm retrieved 70.77% of relevant and partial relevant web pages while Google only 28.79% within top 10 pages. Conclusions/Discussion The researcher's algorithm is more effective than Google's PageRank in terms of the relevance of Web pages to search query for a certain group of people represented by the testing users and queries provided by these users, which was statistically significant at the .01 confidence level.

Competition history

• CSEF 2003 Mathematics & Software · Entry S1229

Resources

• CSEF project summary ↗

Related projects

Closest projects by meaning, across every fair and year in the corpus.

Browse more like this

• All Computational Sciences projects
• CSEF Computational Sciences projects
• Every CSEF 2003 project