home header background

SmarTravel

SMART TRAVEL DIGITAL ECOSYSTEM

The project developed under "Celtic-Next EUROGIA 2020 Call" includes 10 partners from 3 countries.

Contact Us

Project Objective

Providing User-Centered Digital Convenience

It is aimed to facilitate the travel process with personalized and intuitive digital services based on users' reservations, preferences and habits.

Providing a Personalized Travel Experience with Artificial Intelligence

It is aimed to maximize travel satisfaction by analyzing user needs, routines and preferences in depth using artificial intelligence technologies.

Integrating Smart Systems

It is aimed to create a comprehensive tourism network by integrating elements such as smart airports, transportation systems, museums, restaurants and activities to the platform.

Establishing a Secure and Sustainable Tourism Infrastructure

It is aimed to develop a smart tourism infrastructure that complies with sustainability, confidentiality and data security principles, and meets modern needs.

Building a Digital Bridge between the Visitor and the Destination

It is aimed to enable interaction between visitors and destinations through an interface that guides users during their travels, instantly responds to their needs and directs them.

Project Subject

SmarTravel is aimed at providing a secure and integrated digital travel ecosystem powered by artificial intelligence and IoT technologies, meeting the transportation, accommodation and activity needs of users entirely.

sidec

End-to-end Travel Experience

Digital assistance is provided to users throughout their entire travel process by integrating transportation, accommodation and activity services via web and mobile applications.

sidec

Secure and Authorized Information Disclosure

Secure, controlled and authorized data sharing with institutions throughout the journey is made possible with the travel experience that begins with digital identity verification.

sidec

Infrastructure Integrated with Smart Devices

A tourism infrastructure that is compatible and interoperable with smart city devices and services is created through a data model developed in accordance with IoT standards.

What Do We Do?

Developing Multilingual Chatbot

  • An intelligent chatbot specifically designed for travel was developed in Turkish and English.
  • Hotel and tour reservations, changes and cancellations, including payment, can be made via the chatbot.

Providing Speech-based Interaction

  • Models converting Turkish speech to text were developed and integrated into the chatbot, enabling voice interaction.
  • Voice output models were trained and integrated into the system to enable chatbot responses to be voiced in Turkish.

Developing Personalized Recommendation Systems

  • Recommendations for nearby destinations that match the user's interests are provided based on their location and input.
  • Abilities to understand the user's request and generate recommendations based on context were enhanced with natural language processing techniques.

Service Integration and Action Management

  • A structure integrated with the services provided by Setur was established to ensure that requests received via the chatbot are converted into action.
  • User actions were recorded throughout the entire communication process for feedback and service optimization.

Project Output

Developing End-to-end Artificial Intelligence-based Digital Infrastructure for Smart Travel Assistant

An intelligent assistant infrastructure enhanced with GPT and BERT-based artificial intelligence components was developed for travel, enabling voice and text interaction with Turkish language support.

Turkish Natural Language Processing and Speech Models

A Turkish BERT model was trained to ensure accurate comprehension of user intent. Using LightSpeech, a TTS model was developed and a success rate of over 90% was achieved. STT activities were completed above target with a success rate of 91.2%.

Setur Integrations and Action Assistant

Users can easily make hotel and tour reservations, cancellations and changes via the chatbot. A structure directly integrated with Setur services was provided. The assistant is able to comprehend conversations and take the appropriate action.

Management Interfaces and Academic Publications

The assistant management panel provides functions such as intent definition, conversation flow design and user management. The system developed was also disseminated academically through a paper presented at the KDIR 2024 conference. The system yielded both commercial and academic outputs.

Çerezler