Abstract
In this work, we will consider news articles to determine geolocalization of their information and classify their topics on the basis of an available open data source: OpenStreetMap (OSM). We propose a knowledge-based conceptual and computational approach that disambiguates place names (i.e., geo-objects and regions) mentioned in news articles in terms of geographic coordinates. The geo-located news articles are analyzed to identify local topics: we found that the mentioned geo-objects are a good proxy to classify news topics.
Cite
CITATION STYLE
Dashdorj, Z., Khan, M. T., Bozzato, L., & Lee, S. K. (2016). Classification of news by topic using location data. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 10055 LNCS, pp. 305–314). Springer Verlag. https://doi.org/10.1007/978-3-319-50112-3_23
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
