The work is devoted to the development of an expert system in the field of selection of literature on request in natural language, presented in the form of a chatbot that integrates into an existing web application of a social network "VK". The developed system allows you to get a list of the most suitable, as well as similar in content literature, after receiving and processing a text description from the user. The chatbot is designed for personal use by people of all ages. The system allows you to replace the use of standard search filters.
At the moment, there are a large number of applications for searching for specific books, articles, etc. However, the capabilities of search filters of such systems are limited by a standard set of parameters: "Author", "Genre".
To process descriptive user requests, it is necessary to understand how most people make up such appeals and what speech phrases they use. To select the literature suitable for the request, it is necessary to understand what the most significant words can describe or characterize each unit.
The main advantage of the developed system is the use of an alternative search solution to filters. None of the systems on the market provides users with the ability to search without a specific query. Therefore, the creation of a chatbot is relevant today.
Block diagram of the information system:
Datalogical schema of the database:
Class diagram:
| Name | Link |
|---|---|
| .NETFramework, C# | Docs |
| Microsoft SQL Server | Docs |
| VK.NET | Docs |
| Pullenti | Docs |