Trip Nexus

Abstract

The apps for tourism currently available on the market tend to address only one component of the journey effectively, while other components become quite fragmented. If a tourist would like to plan his tour, hire a capable guide, book a package and trust the ratings, he needs to use different apps, which have certain disadvantages. However, the worst issue is related to the reliability of the ratings because they are located in a single place and may be altered or forged without any traces for the tourist to trace. The Final Year Project offers TripNexus, which is an artificial intelligence-enhanced role-based mobile tourism application built using the Flutter programming environment with MongoDB Atlas as the persistence database technology. It has an integration of the Mistral large language model API to provide AI-based itinerary suggestions and landmark recognition features, the Open-Meteo service for acquiring weather information of multiple cities, the Google News RSS feed parser to extract news and event information in each area, and the development of a unique Solidity smart contract in the Polygon Amoy testnet to ensure immutable blockchain-based review storage. The application covers the full cycle of travel services including pre-trip planning, automatic generation of travel itineraries, recognition of landmarks, selection of packages, booking, advisor chat communication, and lastly blockchain submission of travel review after completion of the travel experience. Travel agencies and individual travel advisors control the cataloging of packages, booking processes involving the states of pending, approved, and completed, as well as analyzing reviews of the particular travel packages on the blockchain. This application uses a trust model that writes only completed booking reviews onto the blockchain and keeps all other booking data off the chain in MongoDB. A total of 22 functional requirements and 8 non-functional requirements have been mapped with executed test cases. Tests specifically focused on edge cases have been performed for the blockchain review process flow with respect to the following scenarios: no gas transaction, incorrect contract address, and duplicate submission. Acceptance criteria have been achieved. Hence, using the developed prototype, it has been demonstrated that it is possible to develop a mobile application that includes AI and marketplace functionalities as well as blockchain-based reviews, all accomplished by a small student team.