Development of a Chatbot for the Subject Area Train Schedule for Telegram

• Published in: 2022 International Russian Automation Conference (RusAutoCon) pp. 706 - 710
• Main Authors: Aksyonov, Konstantin, Ziomkovskaya, Polina, Aksyonova, Elena
• Format: Conference Proceeding
• Language: English
• Published: IEEE 04.09.2022
• Subjects: Animals; Automation; Buildings; chatbot; COVID-19; question-and-answer system; Schedules; Semantics; Speech recognition; Telegram; train schedule; Twin
• DOI: 10.1109/RusAutoCon54946.2022.9896293

In this article, using the Twin question-and-answer system, was developed and tested a chatbot for the subject area "Train Schedule" for Telegram. This area is quite promising, because it greatly simplifies the process of communication with customers. During buying tickets, users often have a lot of questions, but not all the information is available on the site. Also, the search process takes a long time, and a person would have to call the train station, where also will be stand in line. This bot has been tested in a variety of situations. The results of testing showed that there are some areas, where the bot responds incorrectly. After analyzing the errors, it was discovered that it was possible to correct these errors and achieve better results.