Background Child injury has been a significant societal problem in many countries and various data-mining methodologies have been employed to extract useful information from various injury data sources. However differences of frameworks and vocabularies used by these different sources hinder the integrated analysis across the different source data. The study aims to develop a conceptual child injury data model, Child Injury Information Ontology (CIIO), and to apply this model to Kids Design Database (KDDB) for capturing useful information for various stakeholders relating to product safety issues.

Methods Firstly, Kids Injury Information Description Framework (KIIDF), a rearranged and non-duplicating set of all attributes of KDDB’s four different databases, is presented. Secondly, an ontology of KIIDF is created using Basic Formal Ontology (BFO) methodology and Ontologies Inverse Engineering (OIE) technique. An ontology editor named “Hozo” is employed for this purpose.

Results Conceptual model of kids injury information with fifty conceptual-level attributes, which are organised into five conceptual categories (host, vector, agent, environment and consequences), and a series of vocabulary standards for the description of those attributes have been produced. Thus, an adequate ontology for child injury information has been obtained. Usefulness of CIIO is presented through the comparative analysis of two product domains, products designed for children and products designed for adults. The conventional analysis indicates that the latter is safer, but not actually safer when CIIO based analysis were made. In order to provide effective protection for children, more care needs to be taken in the design of products for adults.

Conclusions CIIO is a powerful tool for the integration and cross-referencing of various types of child injury information. CIIO would be useful for various data mining works for injury prevention at both social and technical levels.