A Neural Model for Unsupervised Named Entity Classification

The paper describes an unsupervised model for named entity classification into a large number of classes specified by an ontology. The framework is based on an extended model of hierarchical self-organizing maps. As being founded on an unsupervised neural network architecture, the framework can be applied to different languages and domains. Named entities extracted by mining a domain text corpus encode contextual content information, in a distributional vector space. The classification of the extracted named entities into the taxonomy of the given ontology proceeds by associating every named entity to one node of the taxonomy. We experimented the model in the ”Lonely Planet” tourism domain. The taxonomy, the corpus, and the named entities to classify are the ones proposed in the PASCAL ontology learning and population challenge.