Ontology-based knowledge engineering from various agent systems

Many organizations use multi-agent systems as a system solution to deal with complex problems in their organization. However, there are still obstacles to solving semantic aspect problems related to the data and information in different agents' data sources in the multi-agent system. Semantic aspect problems are about data with different names but having similar meanings or data with similar names but having different meanings. In the previous works, the researchers proposed new methodologies, named "Ontology-based Methodology for Multi-Agent Systems (OmMAS)". OmMAS is to develop ontology knowledge to solve semantic aspect problems for multi-agent systems. However, there still needs proof of implementing OmMAS to solve semantic aspect problems in the real-life case study. This research aims to test and implement OmMAS with a real case study to solve semantic aspect problems in the learning agent's domain and demonstrate how we can extract knowledge from different agents' data sources. Furthermore, this research presents the knowledge as a result of ontology development and also analyses and evaluates the advantages and weaknesses of the OmMAS. In this research, there are two main parts presented. The first part is ontology development process to show how to extract the knowledge from learning agents' data sources using OmMAS. It produces ontology knowledge with a semantic relationship scheme to solve semantic aspect problems. The second part is the result of ontology development and discussion to analyse and evaluate the OmMAS. In addition, from the result and discussion section, we mentioned some advantages and weaknesses of OmMAS as an improvement for future work.