Multi-agent neurocognitive algorithm for controlling the reference of speech events of communication of a general artificial intelligence agent in a situation of synchronous multiple dialogues

Basic principles, models and algorithms for controlling the reference of speech messages have been developed based on the creation of a two-circuit model of multi-agent neurocognitive architecture – a superintellecton, which implements the interaction of the subconscious intellecton and the conscious intellecton. Requirements for ontologies of a general artificial intelligence agent, the conditions for their formation, and the functional units of neurocognitive architectures necessary for their effective formation in the training mode are outlined. The results obtained can be used to create speech recognition and understanding systems that are operational when used in noisy environments and situations of multiple synchronous dialogues to improve the quality of recognition using an understanding of the context of situations.