The Classificatim Sense-Mining System

摘要

In this paper, we present the theory from which a methodology has been developed to create Classificatim, a system for sense mining. The system is rule-based and has been conceived with micro-systems in interrelation. Classificatim is composed of 3 systems, Labelgram - a disambiguating parts of speech tagger, Semegram - a sense tagger and a kernel - Classificatim ('verbatim classifier'). The corpus provided by an agro-food industry enterprise has been used to test the system. We explain how the different micro-systems function, what is peculiar to each of them and in what manner our research is original.