The problem of context-aware task sequence planning for independent or weakly related tasks by an intelligent agent has been considered. The principle of matching the task to the context is analyzed. The structure of the context-aware task sequence planning module as part of an intelligent agent and the corresponding algorithm are proposed. In the structure of the planning module, three main groups of blocks are implemented: operations with tasks, operations with the context, determining the relevance of tasks to the context.
In the article the problem of building intelligent agent whose knowledge base core is ontology has been solved. Classification of those systems according to their functioning has been done. For each class appropriate mathematical software has been developed. Intelligent agent models which functioning is based on the ontology has been investigated. The concept of adaptive ontology has been introduced. The model of adaptive ontology is considered as development of the classic model by adding importance weights of the concepts and relations that are stored in the ontology.
An intellectual functioning of the agent as a rational behavior, which is an effort to get the maximum possible benefit from its activities, is considered. For this aim an intelligent agent is searching for new alternatives (methods, algorithms, etc.), use of which gives better results comparing with the reference approach. For such a search the agent uses domain ontology within which it operates.
In the paper the evaluation of the optimization problems that arise during the automatic ontology building on its structures and content. Formed a series of such problems. The methods and algorithms for their solution.
In the paper a method an algorithm and tools for selection of knowledge from a text document are suggested. It is shown that this algorithm has to be multistage and involve hierarchical procedure of concepts recognition of relations, predicates and rules which are introduced into the resulting ontology.