Mitsuru Ishizuka

Relation Extraction from Web Texts with Linguistic and Web Features: a Study on Supervised, Semi-supervised and Unsupervised Relation Extraction Mitsuru Ishizuka

With the explosion of available textual data on the Web, the importance of mapping textual contents into structured representation through automatically harvesting semantic relations from unstructured text has been recognized. In this book, we systematically study two types of relation extraction: relation extraction for linguistic parsing and that for semantic repository construction. For the first type, we investigate the identification of elements from each sentence and their arrangements in a structured format. For the second type, we focus on the extraction of relations between named entities from a local corpus (Wikipedia) while making use of the huge Web corpus. The book demonstrates an interesting view of using respective characteristics of Wikipedia articles and Web corpus, that is to integrate "deep" linguistic analysis on Wikipedia text with redundancy information on the Web. This book can be used as an introductory reading material for students who are interested in Deep Linguistic Processing or Semantic Relation Extraction. It should also be useful as a reference for practitioners in Relational Knowledge Acquisition.