User-Centered Design untuk Membangun Sistem Rekomendasi yang Transparan pada Tema Savoy Platform Woocommerce

Nabila Puspita Dewi, Dr. Ir. Rudy Hartanto, M.T., IPM. ; Dr. Ir Guntur Dharma Putra, S.T., M.Sc.

2026 | Skripsi | TEKNOLOGI INFORMASI

Sistem rekomendasi berbasis Artificial Intelligence (AI) telah menjadi fitur penting dalam e-commerce untuk memberikan rekomendasi produk yang dipersonalisasi. Namun, kurangnya transparansi dalam proses rekomendasi sering kali menurunkan kepercayaan pengguna karena alasan di balik rekomendasi yang diberikan tidak dapat dipahami. Penelitian ini bertujuan merancang antarmuka sistem rekomendasi AI yang transparan pada WooCommerce menggunakan pendekatan User-Centered Design (UCD) guna meningkatkan kepercayaan dan persepsi transparansi pengguna.

Penelitian diawali dengan thematic analysis terhadap ulasan pengguna untuk mengidentifikasi permasalahan pada sistem rekomendasi yang ada. Berdasarkan hasil analisis tersebut, dikembangkan dua prototipe high-fidelity, yaitu prototipe rekomendasi konvensional dan prototipe rekomendasi transparan yang dilengkapi fitur penjelasan. Evaluasi dilakukan menggunakan within-subject A/B testing terhadap 40 partisipan. Pengukuran subjektif meliputi usability, trust, dan perceived transparency, sedangkan pengukuran objektif menggunakan Task Completion Time dan Interaction Count. Analisis data dilakukan menggunakan uji paired sample t-test.

Hasil penelitian menunjukkan bahwa prototipe rekomendasi transparan meningkatkan trust secara signifikan (p = 0.0007) dan perceived transparency (p < 0>usability meningkat, perbedaannya tidak signifikan secara statistik (p = 0.2400). Selain itu, peningkatan waktu penyelesaian tugas dan jumlah interaksi menunjukkan bahwa pengguna lebih banyak mengeksplorasi fitur penjelasan tanpa mengurangi kemudahan penggunaan sistem.

Artificial Intelligence (AI)-based recommendation systems have become an essential feature in e-commerce by providing personalized product suggestions. However, the lack of transparency in many recommendation systems often reduces user trust because users cannot understand the reasoning behind the recommendations. This study aims to design a transparent AI recommendation interface for WooCommerce using the User-Centered Design (UCD) approach to improve user trust and perceived transparency. The research began with a thematic analysis of user reviews to identify usability and transparency issues in existing recommendation systems. Based on the findings, two high-fidelity prototypes were developed: a conventional recommendation interface and a transparent recommendation interface that provides explanation features. Both prototypes were evaluated using within-subject A/B testing involving 40 participants. Subjective measurements included usability, trust, and perceived transparency, while objective measurements consisted of Task Completion Time and Interaction Count. The collected data were analyzed using paired sample t-tests. The results show that the transparent recommendation interface significantly improves user trust (p = 0.0007) and perceived transparency (p < 0 xss=removed>

Kata Kunci : E-commerce, Explainable AI, Pengalaman Pengguna, Sistem Rekomendasi, Transparansi, User-Centered Design