Collaborative conceptualisation processes in the development of lightweight ontologies
Purpose In collaborative settings, such as research and development projects, obtaining the maximum benefit from knowledge management systems depends on the ability of the different partners to understand the conceptualisation underlying the system’s knowledge organisation. This paper aims to show how information/knowledge organisation in a multi-organisation project can be made more effective if the domain experts are involved in the specification of the systems semantic structure. A particular aspect is further studied: the role of conceptual relations in the process of collaborative development of such structures. Design/methodology/approach An action-research approach was adopted, framed by a socio-semantic stance. A collaborative conceptual modelling platform was used to support the members of a research and development project in the process of developing a lightweight ontology aiming at reorganising all the project information in a wiki system. Data collection was carried out by means of participant observation, interviews and a questionnaire. Findings The approach to solve the content organisation problem revealed to be effective both in the result and the process. It resulted in a better-organised system, enabling more efficient project information retrieval. The collaborative development of the lightweight ontology embodied, in fact, a learning process, leading to a shared conceptualisation. The research results point to the importance of the elicitation of conceptual relations for structuring the project’s knowledge. These results are important for the design of methods and tools to support the collaborative development of conceptual models. Originality/value This paper studies the social process leading to a shared conceptualisation, a subject that has not been sufficiently researched. This case study provides evidence about the importance of the early phases of the construction of ontologies, mainly if domain experts are deeply involved, supported by appropriated tools and guided by well-structured processes.