9:00 AM - 9:20 AM
[2L1-01] Location identification using local context and relevant documents
Keywords:information extraction, named entity, location
In this paper, we discuss the task of matching the location name appearing in the document to the geographical location on the real world.In this task, it often happens that one location name points to more than one geographical location.In order to resolve this situation, many previous methods have been proposed for selecting the best candidate from the candidates.For example, a method of outputting a candidate having the largest population was proposed.
In this research, we propose a method to select the best candidate by using local context and relevant documents which are easily obtained and easily preprocessed than the external data necessary for the previous research.Through experiments, the proposed method achieves about 70\% accuracy, and further improvement of accuracy is expected by handling more relevant documents.
In this research, we propose a method to select the best candidate by using local context and relevant documents which are easily obtained and easily preprocessed than the external data necessary for the previous research.Through experiments, the proposed method achieves about 70\% accuracy, and further improvement of accuracy is expected by handling more relevant documents.