Hypotheses Generation Pertaining to Ayurveda Using Automated Vocabulary Generation and Transitive Text Mining

Automated extraction of knowledge from voluminous documents is a vast research area. Text mining is a promising approach for extracting knowledge from unstructured textual documents. The objective of this paper is to mine documents pertaining to Ayurveda, which are retrieved from PubMed into a databank, and find novel transitive associations among biological objects. This paper discusses the extraction of biological objects from the databank using an Automated Vocabulary Discovery (AVD) algorithm. A text-mining process is described for finding transitive (novel) associations among the extracted biological objects. The text mining algorithm, in addition to identifying novel associations (termed hypotheses), also assigns a numerical significance score to them. The expectation is that those with higher score have greater likelihood of being true than those with lower scores. Experimental results as well as their validation results are presented, demonstrating that the method has the potential to predict novel and interesting true associations.