Implementasi Algoritma IBCF Untuk Sistem Rekomendasi Produk Dengan Integrasi Chatbot Whatsapp Otomatis
Building-material stores manage diverse products and transaction records that can be used to identify relationships between products. However, recommendations may still depend on staff experience, while customer questions and orders through WhatsApp require manual responses. This study develops a product recommendation system using Item-Based Collaborative Filtering (IBCF) and integrates it with an automated WhatsApp chatbot through Fonnte. The research uses a Research and Development approach. Data were collected through observation, interviews, and documentation at TB. Rizki Mandiri. Historical transactions are prepared into a user-item matrix, and cosine similarity is applied to measure similarity between products. The recommendation results are ranked to form a Top-N list, while the chatbot uses Fonnte to exchange WhatsApp messages and Ngrok to expose the locally hosted webhook to the internet. The thesis reports that, when Semen is used as the reference product, the system recommends Pasir with a similarity value of 0.67, Cat with 0.63, and Batu with 0.32. Functional tests reported successful operation of login, product and category management, customer data, transaction details, AI features, and chatbot access. These findings show that IBCF can be connected to a WhatsApp-based service workflow to deliver product information and recommendations. Further evaluation with larger transaction data and quantitative recommendation metrics is recommended.
Authors
- Al-Khowarizmi Al-Khowarizmi (ORCID: https://orcid.org/0000-0002-4401-9999)
- Aidil Baqirrahman Nasution
Publication Details
- Journal
- Jurnal Teknologi informasi dan Ilmu Komputer
- Published
- 2026-11-22
- DOI
- https://doi.org/10.65258/jutekom.v2.i4.79
- Primary Topic
- Recommender Systems and Techniques
- Type
- article
- Field-Weighted Citation Impact
- 0.00