Medicinal plant information extraction system—A text mining-based approach

Abstract

In this paper, we have discussed on applying text mining techniques to extract information on health benefits of medicinal plant from text article. The presence of multi-term phrases and complex sentences, i.e., the sentences that include clauses, implicit semantic relations in a text article, has always raised the complexity of information mining process. We have proposed a simple pattern-based semi-supervised approach powered by NLP techniques to deal with these issues. We have evaluated our methodology for a set of web documents on medicinal plants. Performance (in terms of recall) of our method was observed to considerably higher than existing relation extraction methodologies.