Triple extraction using lexical pattern- based syntax model

Abstract

This work proposed a new approach to extract relations and their arguments from natural language text without knowledge base. This approach is mentioned in a paper with the same title which is accepted by ICCSAMA conference in 2016. Using the grammar of English language, it allows detecting sentence based on verb types and phrasal verb in terms of extraction. In addition, this approach is able to extract the properties of objects/entities mentioned in text corpus, which previous works have not yet explored. Experimental result is performed by using various real-world datasets which were used by ClausIE and Ollie, and other text were found in the Internet. The result shows that our method is significant in comparison with ClausIE and Ollie.