Skripsi

SISTEM REKOMENDASI MAKANAN INDONESIA MENGGUNAKAN ALGORITMA LONG SHORT-TERM MEMORY (LSTM) DAN METODE SIMILARITY

---

The vast diversity of Indonesian cuisine often leads to information overload and the "paradox of choice" for users. Existing conventional search systems are unable to understand specific personal preferences, necessitating an intelligent recommendation system. This research aims to (1) design and build a hybrid Indonesian food recommendation system model combining the Long Short-Term Memory (LSTM) algorithm for category classification and similarity methods for recommendation ranking , and (2) analyze the performance comparison between single-layer and double-layer LSTM architectures to determine the most optimal architecture. This study used an Indonesian food recipe dataset from eight categories which underwent text preprocessing and data balancing using Random Undersampling . An LSTM model was trained to classify recipes, then similarity methods such as Cosine Similarity were applied to rank the recommendations. Model performance was evaluated using accuracy, macro precision, macro recall, and macro F1-score metrics. The results showed that the double-layer LSTM architecture (85.91% accuracy) significantly outperformed the single-layer architecture (75.49% accuracy) and successfully overcame the single-layer model's weakness in distinguishing between 'kambing' (goat) and 'sapi' (beef) categories. However, the double-layer model was identified as suffering from over-specialization, achieving 100% accuracy on specific queries but failing on general queries. Both models also consistently struggled to differentiate between the 'tahu' (tofu) and 'telur' (egg) categories. In the ranking stage, the Cosine Similarity method proved to be the most superior and reliable in providing relevant recommendations.

---

Availability

Detail Information

Series Title

-

Call Number

T1896432025

Publisher

Indralaya : Prodi Teknik Informatika, Fakultas Ilmu Komputer Universitas Sriwijaya., 2025

Collation

xiv, 163 hlm.; ilus.; tab.; 29 cm

Language

Indonesia

ISBN/ISSN

-

Classification

006.310 7

Content Type

Text

Media Type

-

Carrier Type

-

Edition

-

Subject(s)

Prodi Teknik Informatika
Data Mining--Algoritma Similirity

Specific Detail Info

-

Statement of Responsibility

MI

Other version/related

File Attachment

Comments

---

You must be logged in to post a comment