Exploiting domain structure for Named Entity Recognition

Named Entity Recognition (NER) is a fundamental task in text mining and natural language understanding. Current approaches to NER (mostly based on supervised learning) perform well on domains similar to the training domain, but they tend to adapt poorly to slightly different domains. We present several strategies for exploiting the domain structure in the training data to learn a more robust named entity recognizer that can perform well on a new domain. First, we propose a simple yet effective way to automatically rank features based on their generalizabilities across domains. We then train a classifier with strong emphasis on the most generalizable features. This emphasis is imposed by putting a rank-based prior on a logistic regression model. We further propose a domain-aware cross validation strategy to help choose an appropriate parameter for the rank-based prior. We evaluated the proposed method with a task of recognizing named entities (genes) in biology text involving three species. The experiment results show that the new domain-aware approach outperforms a state-of-the-art baseline method in adapting to new domains, especially when there is a great difference between the new domain and the training domain.