Automated system of SMM lead generation in Telegram messenger

Abstract

In the article the development of automated system for search of interested users, so called leads, in Telegram messenger environment. While Telegram is not a social network and is strongly different with its interaction mode to ay web-service like blog, image or news board or forum, then the search of motivated target audience is a complex task. It is primarily complex because no recommendation system for content or finding new channels, chats, content sources is provided, the news and posts feed does not exist like in other social media. In current paper the process of development of a tool for searching interested users, created as a Telegram-bot, which interacts with Telegram API to gather the data and with different language tools analyses messages in the chat, helping to find discussions related to required theme. Particularly, to detect users that a potentially interested in specific themes, it is required to analyze the very texts of the discussion and detect the themes, users of the current chat discuss. Specifically for this analysis natural language tools are needed, as well as the tools that allow to process discussion’s context. Bot was created in the following technologies stack: the main programming language is Python, the framework pyTelegramBotAPI is responsible for interaction with Telegram servers via API, the gathered and processed data is stored in a database based on MySQL, language processing is performed in multiple steps, in which natural language processing libraries for Python and AI particularly big language model ChatGPT are involved. The bot gathers and processes information from the chat messages and then provides a report of how many mentions in the administrator defined theme made certain chat users, these users are potential leads. This data helps to build and improve marketing models of goods and services promotion and detect the level of involvement and the degree of interest in current theme