MoonCAB : a Modular Ontology for Computational analysis of Animal Behavior

Computational analysis of animal behavior (CABA) is a modern approach to studying animal behavior using computer techniques. It provides tools for smart farming and analyzes videos mostly with machine learning and deep learning algorithms. The paper aims to integrate ontology in the field of CABA. An ontology is a formal representation of knowledge that can facilitate the integration of data from different sources, allowing experts to better understand and study animal health and behavior. We propose to build a modular ontology based on the Modular Ontology Modeling (MOMo) methodology. The proposed ontology, MoonCAB (Modular ontology for Computational analysis of Animal Behavior), represents the behavior of livestock animals (sheep and goats) in a pasture: the duration of each activity, the meaning of each duration, the season during which these activities take place, etc. Our ontology is composed of 68 classes, 36 properties, 150 individuals, and 8135 axioms. It has been tested by Fact++ reasoner and SPARQL queries.