An Investigation on the use of Ontologies for pattern classification - Study applied to the monitoring of food intake

Abstract

Several tools are developed with the purpose of solving problems and exposing results similar to human reasoning. For this, various artificial intelligence techniques are being implemented to improve these applications. For the poorly structured and high volume data, the ontology presents itself as a technique capable of structuring this data and exposing representative results. In this way, this work describes the use of an ontology as the data classification technique and pattern recognition. The objective is to develop an ontological structure capable of analyzing and classifying the movements and sound signals of the chewing and swallowing process in solids or liquids. To validate the ontology, the tests were performed in real environments. The results obtained, based on the realized experiments, point to the viability of the use of ontologies for the problem in question.