|Adm. - Grad.||2016 -|
|Dir.; Codir.||Stéphane Gagnon|
Semantic Rules Conversion for Multi-Agent Systems
Multi-Agent Systems (MAS) require production rules to automate their monitoring and responses to events. We propose to develop a semantic rules conversion API to help integrate RDF/RDFs/OWL knowledge graphs to an MAS. Rules are expressed with SHACL and ShEx directly linked to RDF triples. They are queried and manipulated in the Elixir language, and converted to production rules in Erlang, both functional languages compiled to the BEAM virtual machine. We test the new rules with an Erlang MAS, SERESYE, where standard and semantic production rules are compared as per their behavior, performance, and scalability in a single MAS use case with varying complexity.