Context Understanding of Cooperative Missions Carried Out by Human-Machine Teams Based on Neurocognitive Models of a Limited Subset of Natural Language

The relevance of the paper is substantiated by the lack of reliable natural language processing systems. This lack hinders the development of intelligent robotics. The article presents the key principles, algorithms and methods of context understanding by an intelligent agent carrying out agricultural tasks. The approach is realized on multi-agent neurocognitive architectures used to model the process of automatic utterance interpretation of a limited subset of natural language. For the interpretation of an input message the intelligent agent needs to identify conditions, actions, attributes and relations in “intelligent agent – environment” system. After that the intelligent agent can interpret the context of the present dialogue and generate utterances to design cooperative behaviour aimed at joint solution to agricultural problems. One common problem to tackle in the field of agriculture is to decrease pesticide intake and reduce chemical load in the environment. At the same time, it is vital to increase effectiveness of plant monitoring and protection to keep a desired level of food production. According to abovementioned it seems to be relevant to design a dialogue control system of joint human-machine teams to interpret aims and current conditions of cooperative missions stated via natural language.