A novel application of data modeling for extracting mathematical ontologies and relationship matrix features from text documents
Holger M. Jaenisch, James W. Handley, Jeffery P. Faucheux, Kenneth Lamkin · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 2005
This paper presents a novel algorithm for extracting Who, Where, When, What, How, and Why information directly from text documents without the use of dictionaries. Our process converts text into mathematical representation and uses purely equation and functional based relationships to make all classification decisions. Applications include requirements matching, hypothesis generation, data mining, document summarization, and document understanding.