GeoSearch: A Geographically-Aware Search Engine

System Overview

As a proof of concept, GeoSearch searches news articles from more than 300 on-line newspapers based in the United States. Off line, GeoSearch estimates the geographical scope of the newspapers based on the distribution of hyperlinks to them. For example:

• The geographical scope of The New York Times is automatically estimated to be the entire United States (see below), which intuitively indicates that this newspaper is generally relevant to users across the country.
• In contrast, the geographical scope of The Stanford Daily is automatically estimated to be mostly the Palo Alto, California area (see below), which intuitively indicates that this newspaper is generally not relevant to users, say, in New York City.

Then, for a query consisting of a list of keywords (e.g., [startups business]) and the US ZIP code of the user's location (e.g., 94043), Geosearch:

1. Uses just the keywords to rank the newspaper articles using a standard, off-the-shelf text search engine called Swish.
2. Filters out all pages coming from newspapers whose geographical scope does not include the user's specified ZIP code.
3. Recomputes the score for each surviving page and returns the pages ranked in the resulting order. A page's new score is a combination of the Swish-generated score for the page and a score related to the geographical scope of the page (see VLDB '00 paper).

Example Geographical Scopes

• Some newspaper geographical scopes, derived automatically from the distribution of hyperlinks to the newspaper homepages:

  The New York Times
  The Digital Missourian
  The Stanford Daily

• Geographical scope of all 300+ newspapers we indexed (might take a minute or two to display)

Papers

• Categorizing Web Queries According to Geographical Locality,  L. Gravano, V. Hatzivassiloglou, and R. Lichtenstein, in Proc. of the 12th ACM Conference on Information and Knowledge Management (CIKM 2003), 2003.
• Computing Geographical Scopes of Web Resources, J. Ding, L. Gravano, and N. Shivakumar, in Proc. of the 26th International Conference on Very Large Data Bases (VLDB '00), 2000.
• Exploiting Geographical Location Information of Web Pages, O. Buyukkokten, J. Cho, H. Garcia-Molina, L. Gravano, and N. Shivakumar, in Proc. of  the ACM SIGMOD Workshop on the Web and Databases (WebDB '99), 1999.

People

• Steve Briskin
• Junyan Ding
• Luis Gravano (contact)
• Vasilis Hatzivassiloglou
• Richard Lichtenstein
• Narayanan Shivakumar

and, at an early stage of the project, Orkut Buyukkokten, Junghoo Cho, and Hector Garcia-Molina.

This material is based upon work supported by the National Science Foundation under Grants No. 9733880 and 9619124. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.