INTEGRATION MODULE OF WEB APPLICATIONS WITH ARTIFICIAL INTELLIGENCE
DOI:
https://doi.org/10.31649/1999-9941-2024-59-1-5-12Keywords:
web application, integration module, artificial intelligence, ChatGpt, multi-tier architecture, .NET, Access TokenAbstract
Abstract. The method of integrating the server part of the web application for organizing a diet and workouts with artificial intelligence ChatGpt was improved in the work, which made it possible to improve the User Experience when using the server part of the web application and increase the speed of processing requests by artificial intelligence and speed up the user's work with artificial intelligence. Modern technologies for the connection of the server part of the web application with the database, the implementation of authorization in order to implement the protection of the server part from intruders and the methods of integrating the web application with artificial intelligence are analyzed. Based on the results of the analysis, a stack of technologies was chosen, in particular, Web-API was used to develop the server part, Entity Framework technology was used to connect the server part of the web application with the database, and authorization was provided using JWT Tokens. The server part of the web application for workouts and diet organization was developed in the C# programming language, taking into account the principles of SOLID object-oriented programming. During the development, a multi-level architecture was used to distribute various functionalities between different modules of the server part. The connection of the server part of the web application with the database for receiving and saving information is configured. An authorization mechanism on JWT-Tokens was implemented, which made it possible to protect the methods for integration with artificial intelligence from attackers and gave the opportunity to only authorized users to use artificial intelligence to build workouts and diets. An MSSQL database that stores personal user data used to build the ChatGpt AI diet and workouts is designed. Database queries are optimized by using the Eager Loading technique and Deferred Execution queries.
References
Introducing ChatGPT and Whisper APIs. [Електронний ресурс]. Режим доступу: https://openai.com/blog/introducing-chatgpt-and-whisper-apis.
Фреймворк Qwik. [Електронний ресурс]. Режим доступу: https://qwik.builder.io/docs/concepts/think-qwik/.
Understanding Multilayer Architecture. [Електронний ресурс]. Режим доступу: https://www.c-sharpcorner.com/UploadFile/1492b1/understanding-multilayered-architecture-in-net/
A Multi-Layer Back-End Application Architecture in .NET Core. [Електронний ресурс]. Режим до-ступу: https://hamzaak.medium.com/a-multi-layer-back-end-application-architecture-in-net-core-c08898f2427e
Introduction to Entity Framework. [Електронний ресурс]. Режим доступу: https://www.partech.nl/nl/publicaties/2020/11/introduction-to-entity-framework
Войцеховська О. В., Городецька О.С., та Рижков А. К., “Аналіз методів авторизації при проек-туванні серверної частини веб-додатку” на Матеріали міжнародної науково-практичної інтер-нет-конференції “Електронні інформаційні ресурси: створення, використання, доступ”, 2023 р. С. 225–227. Режим доступу: https://drive.google.com/file/d/1oVmxS3W_sEQPjes9S9Az
WDaJxDxi6I0X/view.
Authentication and Authorization in Web API. [Електронний ресурс]. Режим доступу: https://dotnettutorials.net/lesson/authentication-and-authorization-in-web-api/
Introduction to authorization in ASP.NET Core. [Електронний ресурс]. Режим доступу: https://learn.microsoft.com/en-us/aspnet/core/security/authorization/introduction?view=aspnetcore-7.0
JSON Web Tokens. [Електронний ресурс]. Режим доступу: https://auth0.com/docs/secure/tokens/json-web-tokens
Prompt engineering. [Електронний ресурс]. Режим доступу: https://platform.openai.com/docs/guides/prompt-engineering.
Embeddings – OpenAI API. [Електронний ресурс]. Режим доступу: https://platform.openai.com/docs/guides/embeddings.
Downloads
-
PDF (Українська)
Downloads: 158