Melola Discovery

← Melola Discovery

Akses terbuka Jurnal SINTA 1

Implementing support vector machine for intelligence allergen classification in the MBG program

International Journal of Education and Learning · Vol. 8 No. 2 (2026) · hlm. 163–176

DOI
10.31763/ijele.v7i2.2362
ISSN
2684-9240
Terbit
2026-08-31
Bahasa
en
Lisensi
creativecommons.org/licenses/by-sa/4.0/

Abstrak

Makan Bergizi Gratis (MBG) is an Indonesian government initiative aimed at improving students' nutritional status and reducing the prevalence of stunting. Currently, the MBG menu planning lacks an automated allergen screening mechanism. To address this limitation, this study introduces a Support Vector Machine (SVM) based allergen classification model for Indonesian food recipes to enhance the safety of MBG menu planning. A dataset of 9,986 Indonesian recipes was compiled via web scraping from the Cookpad website using the keyword “Masakan Indonesia.” Each recipe was annotated according to the 14 major food allergens defined by the European Union, including derivative ingredients. Three classification scenarios were developed: binary classification, four allergen level classes, and fifteen allergen-count classes, each evaluated with four data variations: raw, raw plus Random Oversampling (ROS), augmented, and augmented plus ROS. Four SVM kernels (Linear, Polynomial, Radial Basis Function, and Sigmoid) were assessed using 10-fold cross validation based on accuracy, precision, recall, and F1-score. Experimental results demonstrate that the proposed approach achieved high classification performance across all scenarios. In the most complex fifteen-class scenario, the Polynomial kernel trained on the augmented dataset achieved the highest F1-score of 0.981, while the Linear kernel trained on the augmented plus ROS dataset attained the highest accuracy of 0.999. These results suggest that optimal kernel selection, combined with data balancing and augmentation, can significantly enhance the performance of multiclass allergen classification. Although the proposed model has not yet been implemented in a real MBG setting, it shows strong potential to support AI-assisted allergen screening and decision-making during MBG menu planning prior to food distribution.

Kata kunci

Jaringan kolaborasi para penulis

24 penulis · 66 tautan kepenulisan bersama · 23 lintas organisasi

Arahkan kursor atau ketuk sebuah nama untuk menyorot rekan penulisnya. Klik dua kali untuk membuka halaman penulis.

  • State University of Malang
  • University of South Australia
  • Anadolu University
  • American University of Sharjah
  • Universitas Muhammadiyah Semarang
  • AGH University of Krakow
  • Organisasi berbeda
  • Organisasi sama
  • Afiliasi tidak diketahui
  • Garis lebih tebal = lebih banyak artikel bersama · lingkaran lebih besar = lebih banyak artikel
Tampilkan sebagai tabel
Rekan penulisOrganisasiArtikel bersama
Aji Prasetya WibawaState University of Malang3
Andrew NafalskiUniversity of South Australia2
Muladi MuladiState University of Malang1
Alifah Diantebes AindraState University of Malang1
Didik NurhadiState University of Malang1
Dedi KuswandiState University of Malang1
Roni HerdiantoState University of Malang1
Punaji SetyosarıState University of Malang1
Ilham Mulya Putra PradanaState University of Malang1
Syaad PatmantharaState University of Malang1
Gülsün KurubacakAnadolu University1
Marji MarjiState University of Malang1
Tri KuncoroState University of Malang1
Mazarina DeviState University of Malang1
Ming Foey TengAmerican University of Sharjah1
Felix Andika DwiyantoAGH University of Krakow1
Purnomo PurnomoUniversitas Muhammadiyah Semarang1
Khurin NabilaState University of Malang1

Afiliasi berasal dari OpenAlex; metadata OAI-PMH jurnal tidak memuatnya. Penulis tanpa ID OpenAlex dicocokkan berdasarkan nama, sehingga dua orang bisa tergabung atau satu orang bisa terpisah.

Sumber metadata: OpenAlex API (jurnal SINTA, ADR 0048) https://openalex.org/S4210234871, pertama di-harvest 21 Nov 2026, terakhir terlihat 21 Nov 2026.
Metadata asli (oai_dc)
<oai_dc:dc xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:oai_dc="http://www.openarchives.org/OAI/2.0/oai_dc/"><dc:title xml:lang="en">Implementing support vector machine for intelligence allergen classification in the MBG program</dc:title><dc:creator>Adil Zakaria</dc:creator><dc:creator>Aji Prasetya Wibawa</dc:creator><dc:creator>David Satria Alamsyah</dc:creator><dc:creator>Ahmad ‘Ammar Musyaffa</dc:creator><dc:creator>A W Yulianto</dc:creator><dc:creator>Agung Bella Putra Utama</dc:creator><dc:subject>support vector machine</dc:subject><dc:subject>Makan Bergizi Gratis</dc:subject><dc:subject>food safety</dc:subject><dc:subject>allergen screening</dc:subject><dc:subject>data augmentation</dc:subject><dc:subject>random oversampling</dc:subject><dc:subject>multi-class classification</dc:subject><dc:subject>kernel selection</dc:subject><dc:subject>stunting prevention</dc:subject><dc:subject>web scraping</dc:subject><dc:description xml:lang="en">Makan Bergizi Gratis (MBG) is an Indonesian government initiative aimed at improving students' nutritional status and reducing the prevalence of stunting. Currently, the MBG menu planning lacks an automated allergen screening mechanism. To address this limitation, this study introduces a Support Vector Machine (SVM) based allergen classification model for Indonesian food recipes to enhance the safety of MBG menu planning. A dataset of 9,986 Indonesian recipes was compiled via web scraping from the Cookpad website using the keyword “Masakan Indonesia.” Each recipe was annotated according to the 14 major food allergens defined by the European Union, including derivative ingredients. Three classification scenarios were developed: binary classification, four allergen level classes, and fifteen allergen-count classes, each evaluated with four data variations: raw, raw plus Random Oversampling (ROS), augmented, and augmented plus ROS. Four SVM kernels (Linear, Polynomial, Radial Basis Function, and Sigmoid) were assessed using 10-fold cross validation based on accuracy, precision, recall, and F1-score. Experimental results demonstrate that the proposed approach achieved high classification performance across all scenarios. In the most complex fifteen-class scenario, the Polynomial kernel trained on the augmented dataset achieved the highest F1-score of 0.981, while the Linear kernel trained on the augmented plus ROS dataset attained the highest accuracy of 0.999. These results suggest that optimal kernel selection, combined with data balancing and augmentation, can significantly enhance the performance of multiclass allergen classification. Although the proposed model has not yet been implemented in a real MBG setting, it shows strong potential to support AI-assisted allergen screening and decision-making during MBG menu planning prior to food distribution.</dc:description><dc:date>2026-08-31</dc:date><dc:type>info:eu-repo/semantics/article</dc:type><dc:identifier>https://doi.org/10.31763/ijele.v7i2.2362</dc:identifier><dc:identifier>10.31763/ijele.v7i2.2362</dc:identifier><dc:source>International Journal of Education and Learning; Vol 8 No 2 (2026); 163-176</dc:source><dc:source>2684-9240</dc:source><dc:language>en</dc:language><dc:rights>https://creativecommons.org/licenses/by-sa/4.0/</dc:rights></oai_dc:dc>