Optimalisasi Fitur PRUDENCE pada Tourism Resource Integration Platform: Inovasi Kecerdasan Buatan Generatif untuk Kemandirian Pengambilan Keputusan Berbasis Big Data pada Pengelolaan Desa Wisata Kota Batu

Abstract

The rapid growth of tourism villages in Batu City, Indonesia, has not been matched by an integrated system for managing the large volume of data required for evidence-based decision-making. This study aims to describe and analyze the PRUDENCE (Predictive Resource for User-centered Decision, Evaluation, Navigation, and Consultation Engine) feature embedded in the Tourism Resource Integration Platform (TRIP), a web-based big-data decision-support system developed to strengthen the independence of local governments and tourism-village managers in Batu City. The platform integrates survey data from 1,152 respondents across 384 neighbourhood units (RT) in 24 tourism villages, processed through score aggregation, K-Means clustering, and TOPSIS ranking. PRUDENCE is a generative artificial-intelligence conversational assistant that allows users to query these analytical results in natural language instead of navigating multiple dashboards and tables. A prototype development and testing approach was used, involving functional testing of the assistant's core capabilities and usability testing with tourism-village managers as end users. The results show that PRUDENCE can retrieve village lists, cluster information, priority rankings, and dominant factors, and present them in a simple conversational format, thereby reducing users' dependence on manual data interpretation or external analysts. The feature represents an applied form of AI-powered optimisation that supports industrial and institutional independence in tourism-village management. This study contributes practical evidence on how generative artificial intelligence can be embedded into a big-data platform to enhance evidence-based, self-reliant local decision-making, and suggests further development toward predictive and multi-turn consultation capabilities.