Structured Poi data extraction from Internet news

POI (Point of Interest)database is key resources for GIS (Geographic Information System) application. POI manual gathering is expensive and time consuming. This paper presents a state-of-the-art solution that automatically extracts structured POI data from Internet news. The procedure includes making lexical analysis news Internet, and then identifying time expression, location and organization entities, extracting an event scenario based on POI heuristic features. With POI data extraction, consistency between event and entity, result optimization and filtering with heuristics was taken into account. Open testing with experiment conducted on 1,000 news, the precision is 93.60% and recall is 75.48%. The method within POI oriented event extraction is effective and has been applied in industrial POI collection.