Best behaviour? Ontologies and the formal description of animal behaviour

Abstract

The development of ontologies for describing animal behaviour has proved to be one of the most difficult of all scientific knowledge domains. Ranging from neurological processes to human emotions, the range and scope needed for such ontologies is highly challenging, but if data integration and computational tools such as automated reasoning are to be fully applied in this important area the underlying principles of these ontologies need to be better established and development needs detailed coordination. Whilst the state of scientific knowledge is always paramount in ontology and formal description framework design, this is a particular problem with neurobehavioural ontologies where our understanding of the relationship between behaviour and its underlying biophysical basis is currently in its infancy. In this commentary, we discuss some of the fundamental problems in designing and using behaviour ontologies, and present some of the best developed tools in this domain.

A virtual library for behavioral performance in standard conditions—rodent spontaneous activity in an open field during repeated testing and after treatment with drugs or brain lesions

“…The approach used here in annotations of the raw data objects is a step toward this goal. However, to achieve the goal of searchable linked open datasets requires a more sophisticated and formal schema [ 69–72 ]. The library described here could be of value both as the seed and as the test bed in this endeavor.…”

A knowledge-based system for generating interaction networks from ecological data

“…Limited modelling of behavioural concepts has been undertaken e.g. in the context of human neurobiology [29,30]. In biodiversity and ecosystem informatics (BDEI) aspects of behavioural ecology have been modelled in male jumping-spider courtship behaviour, seaturtle nesting behaviour [31] and the behaviour of social insects [32].…”

Towards a functional classification of behaviour: a taxonomy based on outcomes

“…One alternative approach (related to ethograms) comes from the design of ontologies (or ‘controlled vocabularies’) which aim to systematically annotate classes, instances and properties in domains of knowledge using standardised formal languages, usually for knowledge representation for data management and sharing, and automated reasoning systems (Gruber, 1995). Although these methods have been successful in a wide variety of domains in the mature biosciences, behavioural ontologies still present some difficulties (Gkoutos et al, 2015), not least that these large collaborative efforts require consensus on the appropriate conceptualisation and representation of behaviour. A number of now defunct or abandoned projects (e.g.…”