Institutional Repository of Vilnius University
AI-Powered chatbot for hotel reservation systems /
Abstract
dc:descriptionOverall, this diploma project writes about a chatbot that is AI-based with an addition to an existing hotel reservation system. I initially created a complete reservation system using up-to-date web applications. The aim of this project is to incorporate the smart chat system that will respond to the questions of a hotel and assist the user in reserving a room and providing other automatic assistance in real-time. Generally, the project combines conventional machine learning and contemporary large-language models (LLMs). First, user messages are processed by an intent-classification step which applies both TF-IDF model and linear model. Depending on what the user wants to do (usually, booking, canceling, or information), the classifier sends the chat to the appropriate section of the system. Booking, when a user wishes to make a reservation, the system provides the stepby step slot- extraction to retrieve the checkin and check out dates and number of guests and the type of room. In the case of info, each specific question is handled by a pipeline based on Retrieval-Augmented Generation (RAG): hotel documents are divided into small pieces, converted to vectors using sentence-transformer architectures and indexed by FAISS. During the run time, the chatbot locates and proposes relevant text in accordance with the intent of the user and an LLM, performing through the Groq platform, provides relevant accurate responses with citations. The back-end is coded in FastAPI and modules used to process intent, retrieve, booking logic, session management, and model inference modules. The chatbot is compatible with the current Next.js reservation front-end and allows users to chat in real-time without altering the booking engine. When developing, the emphasis was put on avoiding hallucinations, maintaining the booking process predictable and it is important that response time is quick.
Degree
thesis:*- Grantor dc:publisher
- Institutional Repository of Vilnius University
- Year dc:date
- 2026
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Sukhariev, Tymur,
Subjects
dc:subject × 1Rights
dc:rights- Statement dc:rights
-
- info:eu-repo/semantics/openAccess
- Language dc:language
- eng
Identifiers
dc:identifier.*- Repository record dc:identifier
- https://repository.vu.lt/VU:ELABAETD277688573&prefLang=en_US