«Automation and Informatization of the fuel and energy complex»

Experience in a chatbot development and use to navigate the corporate portal

UDC: 004.896
DOI: -

Authors:

¹ Tyumen Oil Research Center, Tyumen, Russia

Keywords: chatbot, language model, NLU (Natural Language Understanding), natural language processing, intents, entities, Rasa

Annotation:

The article considers the experience of developing a prototype chatbot for navigating through a corporate portal instead of using standard search tools. It is assumed that the introduction of the bot will allow employees to interact more effectively with the information posted on the portal and make the search for data relevant. The article describes the basic concepts related to chatbots: NLU (natural language understanding), entities describing specific data, and statements (any user’s request). The choice of the type of chatbot being developed is being considered. The type of chatbot architecture used is presented, consisting of a natural language processing component, a dialog management component and a trained model. The article discusses the capabilities of neural network generator bots such as ChatGPT, GigaChat, YandexGPT. The criteria for choosing the means for writing a chatbot are given. The main aspects of the development of the chatbot navigation assistant on the corporate portal are given and, the creation is carried out using the Rasa platform.

Bibliography:

1. Merkur'eva E.D., Shaykhlislamova S.V. Chat-boty kak strukturnyy element obshcheniya cheloveka i ob"ekta upravleniya // Avtomatizatsiya i informatizatsiya TEK. – 2023. – № 6(599). – S. 50–56. – DOI: 10.33285/2782-604X-2023-6(599)-50-56
2. Rassel S., Norvig P. Iskusstvennyy intellekt. Sovremennyy podkhod: v 3 t. T 1. Reshenie problem. Znaniya i rassuzhdeniya / per. s angl. A.V. Sleptsova. – 4-e izd. – M.: Dialektika-Vil'yams, 2021. – 704 s.
3. Anran Jiao. An Intelligent Chatbot System Based on Entity Extraction Using RASA NLU and Neural Network // J. of Physics: Conf. Series. – 2020. – Vol. 1487. 2020 4th Int. Conf. on Control Engineering and Artificial Intelligence (CCEAI 2020), Jan. 17–19, 2020, Singapore. – DOI: 10.1088/1742-6596/1487/1/012014
4. GigaChat: dokumentatsiya dlya razrabotchikov. – URL: https://developers.sber.ru/docs/ru/gigachat/overview (data obrashcheniya 03.10.2023).
5. Rasa: Open Source Language Understanding and Dialogue Management / T. Bocklisch, J. Faulkner, N. Pawlowski, A. Nichol // NIPS 2017 Conversational AI workshop. – Long Beach, CA, USA, 2017. – URL: https://arxiv.org/pdf/1712.05181.pdf (data obrashcheniya 03.10.2023).
6. Rasa: Developer Documentation Portal. – URL: https://rasa.com/docs/ (data obrashcheniya 03.10.2023).