txtKnot — Text clustering based concept hierarchy to generalize from different text sources

摘要

Living in the modern technology dependent world, we heavily rely on electronically stored data and information, to come up with sound and timely decisions. Considering the entire information technology world, there exists an unimaginable volume of data which contains a lot of information which is relevant to various kinds of fields. But the problem emerges when we are interested to find out about a particular subject. This is due to its scattered nature of relevant and non-relevant data. Therefore it is fair to say that there exists a critical need for a system which could create an ordered structure that provides a way of modeling the underlying relationships of data elements which will ultimately result in a much easier process of decision making. txtKnot is all about solving the above problem by generating a meaningful hierarchy of concepts from a set of unsorted text documents, thus enabling the visualization of relationships that exist within the set of documents. It consists of four main components namely, Data Extraction Module, Data Pre-processor Module, Text Clustering Module and Concept Hierarchy Generation Module. These four components are integrated together in order to fulfill the main objective of providing an easy to use method of organizing, visualizing, searching and filtering of the huge amount of electronically available unsorted textual data.