Opinion target extraction for short comments
PRICAI'12 Proceedings of the 12th Pacific Rim international conference on Trends in Artificial Intelligence
| Hi-index | 0.00 |
In this paper, we propose a strategy to construct a lexical ontology used for Chinese named entities (NEs) and NE relations (NERs) identification. The lexical information from HowNet (reusable information) and semantic constraints (added information) are semi- automatically integrated into our lexical ontology. In the ontology, addition to the relationships of the concepts in the taxonomy, the non-taxonomic relationships between the concepts have been also built through semantic constraints. The testing results for Chinese named entity and relation identification system have shown that the total average F-measure for six types of NEs and fourteen types of NERs is 83.08% and 70.46% respectively. Therefore, this lexical ontology used for Chinese NE and NER identification is reasonable and effective.