Detecting Points of Interest

Introduction

This study investigates a possible method for retrieving and visualising location specific information applied to the pacmaas project. The semantic lexicon WordNet is used in a combination with a search engine to find Points of Interest (POIs) which are then checked by a GPS lookup algorithm that validates the locations found by comparing them to a set of streets that are in the area of interest that is stored in database. WordNet is used in order to find words from the same semantic class that could also provide useful locations when integrated into the query and to keep the implementation independent of the search term defining the interest. The application in the pacmaas project gives promising results and further efforts may give significant improvements.
The sourcecode of this project is shared under the GPL license on SourceForge.

  1. Introduction
  2. Implementation
  3. Results

Copyright 2008 Thomas Debray