An Application of Fuzzy Inductive Logic Programming for Textual Entailment and Value Mining

The aim of this preliminary report is to give an overview of textual entailment in natural language processing (NLP), to present our approach to research and to explain the possible applications for such a system. Our system presupposes several modules, namely the sentiment analysis module, the anaphora resolution module, the named entity recognition module and the relationship extraction module. State-of-the-art modules will be used but no amount of research will go into this. The research focuses on the main module that extracts background knowledge from the extracted relationships via resolution and inverse resolution (inductive logic programming). The last part focuses on possible economic applications of our research.