Background: Single-Subject Design is used in several areas such as education and biomedicine. However, no suited formal vocabulary exists for annotating the detailed configuration and the results of this type of research studies with the appropriate granularity for looking for information about them. Therefore, the search for those study designs relies heavily on a syntactical search on the abstract, keywords or full text of the publications about the study, which entails some limitations. Objective: To present SSDOnt, a specific purpose ontology for describing and annotating single-subject design studies, so that complex questions can be asked about them afterwards. Methods: The ontology was developed following the NeOn methodology. Once the requirements of the ontology were defined, a formal model was described in a Description Logic and later implemented in the ontology language OWL 2 DL. Results: We show how the ontology provides a reference model with a suitable terminology for the annotation and searching of single-subject design studies and their main components, such as the phases, the intervention types, the outcomes and the results. Some mappings with terms of related ontologies have been established. We show as proof-of-concept that classes in the ontology can be easily extended to annotate more precise information about specific interventions and outcomes such as those related to autism. Moreover, we provide examples of some types of queries that can be posed to the ontology. Conclusions: SSDOnt has achieved the purpose of covering the descriptions of the domain of single-subject research studies.
翻译:背景:单受试者设计广泛应用于教育和生物医学等领域。然而,目前缺乏适用于以适当粒度标注此类研究详细配置和结果的规范化术语体系。因此,针对该类研究设计的检索主要依赖于对相关文献摘要、关键词或全文的语法层面搜索,这导致了一定的局限性。目的:提出SSDOnt——一个用于描述和标注单受试者设计研究的特定领域本体,以支持后续针对其内容的复杂查询。方法:采用NeOn方法论开发该本体。在明确本体需求后,基于描述逻辑构建形式化模型,并最终以本体语言OWL 2 DL实现。结果:我们展示了该本体如何为标注和检索单受试者设计研究及其核心要素(如阶段、干预类型、结局指标和研究结果)提供具有恰当术语体系的参考模型。已建立与相关本体的术语映射关系。概念验证表明,本体中的类可方便地扩展为特定干预和结局指标(如自闭症相关研究)的精确标注。此外,我们还提供了可向本体提交的多种查询类型示例。结论:SSDOnt成功实现了对单受试者研究领域描述的覆盖目标。