Artificial Intelligence-Powered Smart Itinerary Generator for Sri Lankan Tourism: Multi Objective Optimization for the Free Independent Traveler Segment

Abstract

The paper is addressing the significant issue of discontinuity of travel information and free independent is faced with a complex of planning. There is the presence of travelers who are approximated to constitute an estimated 80 percent of the tourists visiting Sri Lanka. Although there are several local artificial intelligence systems that support text, most do not offer customized, integrated, and safe itinerary generation, especially in cases where they combine multi-modal transport and follow ethical principles. To address these flaws, the proposed system is an intelligent web and mobile-based system called TravelPlan. The unique hybrid architecture used by TravelPlan: the user preferences are analyzed in the form of the machine learning models, and a genetic algorithm is the central itinerary optimization engine. The most important contribution is a stringent multi-criteria fitness function that streamlines the itinerary by reducing the travelling time and cost as well as maximizing user interest, route efficiency, and most importantly, a safety cultural weighting score developed by the company. Methodology describes the process of putting fragmented transport, accommodation, and point-of-interest information together into one geographic information object. A comparison between the performance of the system and the performance of traditional manual planning is provided through an evaluation plan using both technical metrics including latency and budget adherence and the system usability scale to indicate that TravelPlan will be able to foster efficient, responsible, and high-satisfaction independent tourism in Sri Lanka