<![CDATA[The rapid growth of digital technology and the culinary industry in Tangerang presents new challenges for the community in finding suitable food options. Many users experience difficulties selecting restaurants or dishes that match their individual preferences, creating a need for a more intelligent and adaptive recommendation system. To address this problem, this study develops an AI-based culinary recommendation model that integrates CBF (Content-Based Filtering) and CF (Collaborative Filtering) approaches. The proposed hybrid system combines user behavior patterns with food attributes such as dish type, main ingredients, taste, and price. Data were collected from 90 respondents in Tangerang through questionnaires and interviews, containing user reviews, ratings, and restaurant information. Several hybrid strategies were implemented, including weighted, switching, feature combination, and cascade hybrid methods. Evaluation of system performance used Precision (73%), Recall (76.9%), MAE (0.49), and MSE (0.256). In addition, UAT(User Acceptance Testing) was applied to ensure that the developed system meets functional, usability, and business workflow requirements. The UAT result of 81.86% indicates that the system performs well, is easy to use, and aligns with user expectations. The resulting AI-driven recommendation model successfully provides more accurate, relevant, and personalized culinary suggestions for users. This research contributes to advancing the development of recommendation systems by addressing the limitations of standalone CBF and CF techniques. The proposed hybrid framework offers a practical solution to enhance user experience and strengthen the digital ecosystem of the culinary industry in Tangerang City.]]>
