An Ontology for the Common Data Format on Football Match Data

Abstract

Artificial intelligence (AI) applications in sports, particularly for football (soccer), have been growing in recent years, e.g., for player recruitment, performance monitoring, and selection. To support such applications, the availability of an integrated, high-quality dataset is crucial to ensure accurate results. This aspect is especially vital due to the heterogeneity in data acquired by various stakeholders, e.g., companies and football clubs. Catering to such demand, a recent work proposed a common data format (CDF) schema for football match data to ensure the provided data is precise, sufficiently contextualized, and complete to enable typical downstream analysis tasks. This paper reports on an initial effort to create the Football Common Data Format (FCDF) ontology as a schema for the RDF serialization of the CDF core concepts, focusing on streamlining concepts, properties, and attributes. The FCDF ontology aims to provide a formal, shared conceptualisation of CDF to promote using ontology and KGs for AI applications in football.