🎭 Thea: NLP-Driven Theater Ticketing Application

You can watch the Demo Video here and the Documentation here.

Overview

Thea is a cross-platform mobile application designed to simplify theater ticket booking through a natural language interface.
The chatbot enhances the user experience by providing essential information about shows, assisting with ticket bookings, and managing other requests in an intuitive and user-friendly manner.
The app prioritizes accessibility for both tech-savvy and elderly users, integrating both manual menus and an NLP-powered chatbot for a seamless user experience.

---

Key Features

• Real-time Show Information:
  Instantly ask the chatbot for current performance details, including which hall a show is in and afternoon/evening schedules.

• Conversational Ticket Booking:
  The chatbot walks users step-by-step through ticket reservations—choose your show, select seats, confirm, or cancel before finalizing.

• Easy Ticket Cancellation:
  Cancel booked tickets quickly and intuitively via chatbot conversation, with confirmation to prevent mistakes.

• Complaints and Feedback:
  Submit complaints or feedback directly through the chatbot. Issues can be escalated to a human agent for further assistance.

• Accessibility Support:
  The app supports accessible seating (e.g., people with mobility disabilities) and offers supertitles for certain shows to enhance the experience for everyone.

• Accessibility Settings:
  Personalize your experience with options for dark mode, adjustable font sizes, and larger buttons, ensuring comfortable use for all users, including those with visual impairments.

• Manual & Guided Actions:
  Traditional menu navigation is available alongside the NLP-powered chatbot, making the app friendly for both tech-savvy and less experienced users.

• Robust Error Handling:
  Confirmation dialogs prevent unwanted actions. After repeated misunderstandings, the system can seamlessly escalate to a human agent.

• Support for Elderly and Inclusive Design:
  Large buttons, readable fonts, simple navigation, and optional voice interaction ensure the app is usable by people of all ages and abilities.

• Multi-Stage Booking & Management:
  Supports viewing plays, selecting dates/times, choosing seats, making payments, and managing your tickets all in one place.

• Intelligent Backend:
  LLM-based chatbot (LLaMA 3.2), RAG pipeline (LangChain + FAISS), and a secure Flask API power the conversational and ticketing logic.

---

Tech Stack

• Flutter (Dart):
  Modern, cross-platform framework for building beautiful, responsive mobile apps, ensuring seamless performance and a consistent user experience on both iOS and Android devices.

• Flask (Python):
  Lightweight, high-performance backend framework that handles API requests, ticket management, and serves as the bridge between the mobile frontend and AI-driven chatbot services.

• Large Language Model (LLM):
  Advanced NLP technology (LLaMA 3.2) powers the chatbot, enabling accurate understanding of user intent, natural conversations, and intelligent assistance throughout the app.

• Retrieval-Augmented Generation (RAG):
  Integrates with LangChain and FAISS to empower the chatbot with real-time access to theater information and context-aware responses, delivering up-to-date answers based on backend data.

• FAISS:
  High-speed vector database for efficient retrieval of relevant information, ensuring the chatbot delivers accurate, context-rich answers to user queries.

• LangChain:
  Orchestrates the conversation flow and connects the LLM with the backend, enabling flexible, robust, and reliable information retrieval in real time.

---

Why This Stack?

This tech stack enables scalability, real-time performance, and a premium user experience. Leveraging the latest advancements in mobile development and conversational AI, Thea delivers an innovative, accessible, and reliable solution for theater-goers of all ages.

System Architecture

• Frontend: Mobile app built according to HCI best practices (Figma/Android Studio)
• Backend:
  • Flask API for ticketing logic
  • LLM chatbot (LLaMA 3.2)
  • LangChain & FAISS for Retrieval-Augmented Generation (RAG)
• Chatbot:
  • Understands free-form requests
  • Error handling and escalation
  • Complaint submission flow
  • Integrated fallback to human support

---

Prototypes

See Paper Prototypes and Figma Prototypes.

---

📁 Installation & Usage

1. Clone this repo:

   git clone https://github.com/MariaSchoinaki/Thea.git
   cd Thea
   

2. Navigate to the AI script directory:

   cd lib\chatbot

3. Start the AI server:

   • Open the llama_chat_with_rag_faiss.ipynb notebook.
   • Run all cells to instantiate the server.

4. Launch the mobile app:

   • Connect your mobile phone to your development environment, or use an emulator.
   • From the project root directory, run:

     flutter run

Developers

Maria Schoinaki, BSc Student
Department of Informatics, Athens University of Economics and Business
p3210191@aueb.gr

Eleni Kechrioti, BSc Student
Department of Informatics, Athens University of Economics and Business
p3210078@aueb.gr

Giorgos Isoufai, BSc Student
Department of Informatics, Athens University of Economics and Business
p3210056@aueb.gr