Creating a chatbot using text information processing methods

Abstract

The article considers the relevance and applied aspects of the use of chatbots. The structural and functional system of the chatbot is analyzed. A comparative description of the platforms for creating chatbots has been made. The functional blocks are based on the example of the Telegram social network, a prototype of the system. As part of the chatbot development, the necessary and sufficient condition for working with text is described - the use of various text analysis methods as an input element for interacting with the bot and its training. The article discusses a number of solutions used for text analysis and creation of text data analysis models: lemmatization methods, text vectorization, various machine learning models. The article focuses on the consideration of the methodology of text processing in different formats and using different technologies, which ensures the scalability and versatility of the proposed technology and the overall effectiveness of the future work of the chatbot.The article will be of interest to programmers, text analysts and anyone interested in working with text and creating systems for working with text information. A chatbot has been created in the Telegram messenger in the Python programming language that automates the work of a customer service specialist.