GeoSearcher: Geo spatial ranking of search engine results

The World Wide Web is a rich source of information in a variety of topics. Nevertheless, finding the desired information requires multiple queries and sometimes multiple search methods. Geographic characteristics are important aspect in many web searches, such as finding entities in a specific region or obtaining the closest service to an area. Linking geographic locations to web searches is an emerging area of study that aims to improve search engines and to provide a method for data mining. This work provides a search system called GeoSearcher that follows the described method to improve finding information in the web. The prototype system uses a search engine to answer a search query, obtains sites returned by the search engine, assigns a geographic location to each site and ranks those sites according to the geographic distances from a predefined point. By combining many heuristics to map sites, GeoSearcher was able to map an average of 80% of returned sites to their physical locations. GeoSearcher prototype system provides a step beyond common search engines that takes into consideration users and sites physical locations. From examples illustrated in the thesis, we demonstrate an instance of the ability to combine geographic information with regular keyword searching in a useful manner and identify the opportunities to many future directions.