Geotagging for Malay documents using Name-Entity Recognition Approach

Abstract

There are many tourist places in Malaysia, and still many to be discovered. The Internet is an agent to find web documents describing a place or a location. The precise location or place is always connected to the geotagging process. However, a limited geotagging approach is applied in Malay documents. Hence, a geotagging application is developed using the name-place entity recognition approach. A geotagging algorithm is developed for tagging geographic information in Malay documents. The algorithm performs word filtering and name-place extraction from both the MyGazetteer resources and tourism web resources. An application is built to demonstrate the tagging of geographic information onto each web document. The geotagging prototype (MyGeo-NER) allows users to edit tags of existing web documents, add new ones, and search for documents containing the names of places or locations they have entered.