Search academic texts

Information × Registration Number 0224U000213, 0123U103228 , R & D reports Title Develop methods and tools using a virtual dialog assistant for digital agriculture and environmental and medical projects popup.stage_title Head Litvin Anna A., молодший науковий співробітник Registration Date 03-01-2024 Organization V.M. Glushkov Institute of Cybernetics of the National Academy of Sciences of Ukraine popup.description2 The following results were obtained during the reporting period: 1) A new method and its software implementation is developed to create a deep semantically structured ontology using plain text in natural language as input data (without regular structure or any previous tags and markup). 2) A comprehensive approach to the development of natural language dialog systems using a knowledge base based on ontological graphs has been created. 3) A comprehensive methodology of using an ontology-driven system of structured instructions in interaction with ChatGPT has been studied 4) The features of the wireless sensor network for express diagnostics of the state of plants, monitoring of the environment and the state of human health were considered in the context of using the collected data as a source of information for a dialog assistant. Creating a dialog system that can be "taught" using natural language texts without regular structures or prior markup is an important and urgent task. The automation of this process will significantly help to work with a significant amount of information coming from wireless systems for express diagnostics of the condition of plants, monitoring of the environment and the state of human health. Such a system can help users find answers to their questions in the form of relevant contexts extracted from texts, or even draw conclusions based on the semantic data they present, obtained by automated text analysis; A method has been formalized that involves meta-learning to create and configure structured prompts and an ontology of contexts, as well as a sequence of operations to detect intent, identify named entities, select contexts, and form a prompt instruction that allows ChatGPT to provide certain information by providing selected contexts that can improve his knowledge in specific subject areas and improve his productivity when working with data that is presented in inflected language. Product Description popup.authors Kedych Anna V. popup.nrat_date 2024-01-03 Close