Febrian, Sandy (2026) PENGEMBANGAN PROTOTIPE SISTEM PENCARIAN KOLEKSI BERBASIS SEMANTIK PADA WEBSITE ONLINE PUBLIC ACCESS CATALOG (OPAC) PERPUSTAKAAN STMIK WIDYA CIPTA DHARMA. S1 Teknik Informatika thesis, STMIK Widya Cipta Dharma.
![[img]](http://repository.wicida.ac.id/style/images/fileicons/text.png) |
Text
2143060-S1-Teknik Informatika.pdf Restricted to Repository staff only Download (4MB) | Request a copy |
|
Text
2143060-S1-Jurnal.pdf Download (1MB) |
Abstract
Sistem Online Public Access Catalog (OPAC) pada website Perpustakaan STMIK Widya Cipta Dharma masih menggunakan pencocokan kata kunci secara literal dalam proses pencarian informasi. Pendekatan ini menyebabkan sistem belum mampu memahami konteks pencarian pengguna dan belum optimal dalam menangani kesalahan penulisan kata kunci, sehingga berdampak pada rendahnya efektivitas pencarian informasi. Oleh karena itu, penelitian ini bertujuan untuk mengembangkan prototipe sistem pencarian OPAC berbasis semantik guna meningkatkan relevansi hasil pencarian. Pengembangan sistem dilakukan menggunakan metode Prototyping yang meliputi tahapan komunikasi awal; perancangan dan desain cepat; pembangunan prototipe; penyebaran, pengiriman, dan umpan balik; serta penyempurnaan prototipe. Pencarian semantik diterapkan dengan memanfaatkan model Sentence-Bidirectional Encoder Representations from Transformers (S-BERT) untuk mengubah kata kunci menjadi representasi vektor semantik, Facebook AI Similarity Search (FAISS) sebagai mekanisme pencarian vektor, serta algoritma Cosine Similarity untuk mengukur tingkat kemiripan antar data. Penanganan kesalahan penulisan kata kunci dilakukan menggunakan RapidFuzz dengan algoritma Levenshtein Distance, serta dilengkapi fitur saran pencarian. Sistem juga menyediakan fitur rekomendasi koleksi menggunakan pendekatan Content-Based Filtering berdasarkan riwayat pencarian pengguna. Hasil pengujian menunjukkan bahwa sistem mampu menampilkan hasil pencarian yang relevan meskipun kata kunci tidak sama persis dengan data koleksi atau mengandung kesalahan penulisan. Evaluasi pengguna menggunakan kuesioner skala Likert lima tingkat dengan metode Mean analysis menunjukkan bahwa relevansi hasil pencarian memperoleh nilai rata-rata 4,0 (baik), penanganan kesalahan penulisan 4,4 (sangat baik), kemudahan penggunaan 4,2 (baik), manfaat fitur rekomendasi 4,1 (baik), serta kelayakan sistem secara keseluruhan 4,4 (sangat baik). Berdasarkan hasil tersebut, prototipe sistem OPAC berbasis semantik dinyatakan efektif dan dapat diterima dengan baik oleh pengguna. ============================================================ The Online Public Access Catalog (OPAC) system on the website of the STMIK Widya Cipta Dharma Library still relies on literal keyword matching in the information retrieval process. This approach causes the system to be unable to effectively understand the context of user queries and to optimally handle spelling errors, which in turn reduces the effectiveness of information retrieval. Therefore, this study aims to develop a semantic-based OPAC search system prototype to improve the relevance of search results. The system was developed using the Prototyping method, which consists of the stages of initial communication; rapid planning and design; prototype development; deployment, delivery, and feedback; as well as prototype refinement. Semantic search was implemented by utilizing the Sentence-Bidirectional Encoder Representations from Transformers (S-BERT) model to transform search keywords into semantic vector representations. Facebook AI Similarity Search (FAISS) was employed as the vector search mechanism, while Cosine Similarity was used to measure the level of similarity between data. Spelling error handling was performed using RapidFuzz with the Levenshtein Distance algorithm and was complemented by search suggestion. In addition, the system provides a collection recommendation feature using a Content-Based Filtering approach based on users’ search history. The testing results indicate that the system is able to present relevant search results even when the entered keywords do not exactly match the collection data or contain spelling errors. User evaluation was conducted using a five-point Likert scale questionnaire analyzed with the Mean analysis method. The evaluation results show an average score of 4.0 (good) for search result relevance, 4.4 (very good) for spelling error handling, 4.2 (good) for ease of use, 4.1 (good) for the usefulness of the recommendation feature, and 4.4 (very good) for overall system feasibility. Based on these results, the developed semantic-based OPAC system prototype is considered effective and well accepted by users.
| Item Type: | Thesis (S1 Teknik Informatika) |
|---|---|
| Additional Information: | Pembimbing 1: Amelia Yusnita, S.Kom., M.Kom Pembimbing 2: Ita Arfyanti, S.Kom., M.M |
| Uncontrolled Keywords: | OPAC, Pencarian Semantik, Sistem Temu Kembali Informasi, Penanganan Kesalahan Penulisan, Rekomendasi Koleksi |
| Subjects: | Q Science > QA Mathematics > QA75 Electronic computers. Computer science |
| Divisions: | Teknik Informatika |
| Depositing User: | Mr Sandy Febrian |
| Date Deposited: | 30 Jan 2026 06:42 |
| Last Modified: | 30 Jan 2026 06:42 |
| URI: | http://repository.wicida.ac.id/id/eprint/6293 |
Actions (login required)
 |
View Item |
