Sistem Rekomendasi Menu Makanan Berbasis Content-Based Filtering dan XGBoost untuk Optimasi Kebutuhan Nutrisi Personal
DOI:
https://doi.org/10.36312/r3ne6g79Keywords:
Content-Based Filtering; kebutuhan nutrisi; rekomendasi menu makanan; sistem rekomendasi; XGBoostAbstract
Pemilihan menu makanan yang sesuai dengan kebutuhan nutrisi personal masih menjadi persoalan praktis, terutama ketika pengguna harus menyeimbangkan preferensi rasa, jenis bahan, metode memasak, dan target makronutrisi. Penelitian ini bertujuan merancang dan mengevaluasi sistem rekomendasi menu makanan berbasis aplikasi Android yang mengintegrasikan Content-Based Filtering dan XGBoost untuk optimasi kebutuhan nutrisi. Sistem menghitung kebutuhan energi pengguna menggunakan rumus Revised Harris-Benedict berdasarkan usia, jenis kelamin, berat badan, tinggi badan, dan tingkat aktivitas fisik. Content-Based Filtering digunakan untuk menyaring kandidat lauk berdasarkan jenis bahan utama, cita rasa, dan metode memasak menggunakan cosine similarity berbobot, sedangkan XGBoost Regressor digunakan untuk menilai kesesuaian nutrisi setiap kombinasi menu nasi, lauk, sayur, dan buah. Model dilatih menggunakan 500 profil pengguna simulasi dan 1.000 kombinasi menu sehingga terbentuk 500.000 data pelatihan. Hasil pengujian menunjukkan akurasi perhitungan nutrisi sebesar 99,96% terhadap perhitungan manual dengan rata-rata selisih 0,24 kkal. Model XGBoost menghasilkan MAE 0,0285, RMSE 0,0360, R² 0,9465, Spearman Rank Correlation 0,9654, Top-3 Accuracy 1,0000, dan NDCG@3 0,9922. Pengujian Black Box mencapai keberhasilan 98,6%, sedangkan pengujian pengguna memperoleh skor 4,47 dari skala 5. Temuan ini menunjukkan bahwa integrasi penyaringan berbasis preferensi dan penilaian nutrisi berbasis machine learning mampu menghasilkan rekomendasi menu yang adaptif, terukur, dan relevan bagi pengguna dewasa sehat.
Selecting meals that match personal nutritional needs remains a practical challenge because users must balance food preference, ingredient type, cooking method, and macronutrient targets simultaneously. This study aims to design and evaluate an Android-based food menu recommendation system that integrates Content-Based Filtering and XGBoost for nutritional needs optimization. The system calculates users' energy requirements using the Revised Harris-Benedict equation based on age, sex, body weight, height, and physical activity level. Content-Based Filtering filters side-dish candidates using weighted cosine similarity based on main ingredient type, taste, and cooking method, while XGBoost Regressor evaluates the nutritional suitability of each complete menu combination consisting of rice, side dish, vegetables, and fruit. The model was trained using 500 simulated user profiles and 1,000 menu combinations, resulting in 500,000 training records. The test results show that the nutritional calculation achieved 99.96% accuracy compared with manual calculation, with an average difference of 0.24 kcal. The XGBoost model obtained an MAE of 0.0285, RMSE of 0.0360, R² of 0.9465, Spearman Rank Correlation of 0.9654, Top-3 Accuracy of 1.0000, and NDCG@3 of 0.9922. Black Box testing reached a 98.6% success rate, while user testing achieved an average score of 4.47 out of 5. These findings indicate that integrating preference-based filtering and machine learning-based nutritional scoring can produce adaptive, measurable, and relevant menu recommendations for healthy adult users.
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Erpiana Erpiana, Syafrijon Syafrijon, Yeka Hendriyani, Vera Irma Delianti

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
Authors who publish with Journal of Authentic Research agree to the following terms:
- For all articles published in Journal of Authentic Research, copyright is retained by the authors. Authors give permission to the publisher to announce the work with conditions. When the manuscript is accepted for publication, the authors agrees to implement a non-exclusive transfer of publishing rights to the journals.
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution-ShareAlike 4.0 International License that allows others to share the work with an acknowledgment of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgment of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.