Text-to-Onto Miner: A Concept Driven and Interval Controlled Ontology Builder

Abstract

This is a work-in-progress research paper that describes a novel architecture of a prototype, Text-to-Onto Miner, which is aimed to aid human experts in analyzing huge collection of internet text content by soliciting text from various sources such as blog and publicly accessible web sites. The Text to-Onto Miner or Text-to-Onto is a tool designed and implemented after detail understanding the needs of some fund agencies where existing tools may not be able to fulfill some functional specifications specified by them. The design of the Text-to-Onto Miner uses OWL as the container to hold relationships and key concepts. The design of the prototype focuses on two criteria: concept driven text mining and interval base analysis. These criteria will yield analysis that is interval controlled key concept analysis which can be visually presented their change over time. In the testing and evaluation stage, the prototype has shown that it has achieved goals defined in the project. Some future improvements are proposed to further enhance the prototype to increase the efficiency and to reduce the redundancy of some modules in the prototype.