This paper describes a Machine Learning (ML) approach for extracting named entities and disambiguating the location of tweets based on those named entities and related content. We conducted experiments with tweets (e.g., about potholes), and found significant improvement in disambiguating tweet locations using a ML algorithm along with the Stanford NER. Adding state information predicted by our classifiers increases the possibility to find the state-level geo-location unambiguously by up to 80%.
- Date of publication:
- June 21, 2015
- Joint Conference on Digital Libraries