Abstract: A system, method and computer program product for disambiguating one or more entity mentions in one or more documents. The method facilitates the simultaneous linking entity mentions in a document based on convolution neural networks and recurrent neural networks that model both the local and global features for entity linking. The framework uses the capacity of convolution neural networks to induce the underlying representations for local contexts and the advantage of recurrent neural networks to adaptively compress variable length sequences of predictions for global constraints. The RNN functions to accumulate information about the previous entity mentions and/or target entities, and provide them as the global constraints for the linking process of a current entity mention.

Abstract: A system, method and computer program product for disambiguating one or more entity mentions in one or more documents. The method facilitates the simultaneous linking entity mentions in a document based on convolution neural networks and recurrent neural networks that model both the local and global features for entity linking. The framework uses the capacity of convolution neural networks to induce the underlying representations for local contexts and the advantage of recurrent neural networks to adaptively compress variable length sequences of predictions for global constraints. The RNN functions to accumulate information about the previous entity mentions and/or target entities, and provide them as the global constraints for the linking process of a current entity mention.

Abstract: A system, method and computer program product for disambiguating one or more entity mentions in one or more documents. The method facilitates the simultaneous linking entity mentions in a document based on convolution neural networks and recurrent neural networks that model both the local and global features for entity linking. The framework uses the capacity of convolution neural networks to induce the underlying representations for local contexts and the advantage of recurrent neural networks to adaptively compress variable length sequences of predictions for global constraints. The RNN functions to accumulate information about the previous entity mentions and/or target entities, and provide them as the global constraints for the linking process of a current entity mention.

Abstract: An input entity pair including a first entity and a second entity is received. The first entity and the second entity are associated by a particular relation. A first set of statements containing the first entity and a second set of statements containing the second entity is are received from a corpus. A first set of discriminative words in the first set of statements and a second set of discriminative words in the second set of statements are identified. Perceptions in the first set of discriminative words and the second set of discriminative words are identified. A predetermined number of ranked statements is retrieved from the corpus using the identified perceptions as a query. Candidate entity pairs are extracted from the predetermined number statements. The candidate entity pairs have a relation therebetween analogous to the relation between the first entity and the second entity.

Abstract: An input entity pair including a first entity and a second entity is received. The first entity and the second entity are associated by a particular relation. A first set of statements containing the first entity and a second set of statements containing the second entity is are received from a corpus. A first set of discriminative words in the first set of statements and a second set of discriminative words in the second set of statements are identified. Perceptions in the first set of discriminative words and the second set of discriminative words are identified. A predetermined number of ranked statements is retrieved from the corpus using the identified perceptions as a query. Candidate entity pairs are extracted from the predetermined number statements. The candidate entity pairs have a relation therebetween analogous to the relation between the first entity and the second entity.

Abstract: A system, method and computer program product for disambiguating one or more entity mentions in one or more documents. The method facilitates the simultaneous linking entity mentions in a document based on convolution neural networks and recurrent neural networks that model both the local and global features for entity linking. The framework uses the capacity of convolution neural networks to induce the underlying representations for local contexts and the advantage of recurrent neural networks to adaptively compress variable length sequences of predictions for global constraints. The RNN functions to accumulate information about the previous entity mentions and/or target entities, and provide them as the global constraints for the linking process of a current entity mention.