A semantic representation model based on multi-graph
This paper describes the work on automated knowledge representation that extracts information from an input sentence. A new semantic representation model to implement knowledge representation based on multi-graph is proposed in this paper. With lexical processing, sentence will be compared with the templates knowledge base, and then append the knowledge extracted to the semantic representation model. The model contains three semantic meaning of object, such as attribute, relationship and events. Sentence with deep semantic meaning could be expressed by this semantic representation model. With the experimental corpus the accuracy rate of semantic representation is 63.5%, and accuracy rate of question answering is 65.0% with the semantic representation model.
Author's Name: Chen, Y., Fan, X.
Volume: Volume 7
Issues: Issue 6
Keywords: Multi-graph, Natural language processing, Semantic representation