Pengembangan Sistem Pengelolaan DUPAK Berbasis Web dengan Scrapper Otomatis Menggunakan Metode DSR dengan Pemodelan UML

Proses pengajuan dan penilaian Daftar Usulan Penetapan Angka Kredit (DUPAK) di Fakultas Teknik Universitas Muhammadiyah Surabaya saat ini masih menghadapi kendala signifikan terkait efisiensi waktu, validitas data publikasi, dan kompleksitas perhitungan poin yang harus sesuai dengan Pedoman Operasional Penilaian Angka Kredit (PO PAK) terbaru. Metode manual yang berjalan menyebabkan tingginya risiko redundansi data, kesalahan input, serta lamanya proses verifikasi berkas. Penelitian ini bertujuan mengembangkan sistem pengelolaan DUPAK berbasis web yang mengintegrasikan teknik web crawling untuk otomatisasi data publikasi dan workflow automation menggunakan n8n, yang dirancang menggunakan pendekatan Design Science Research (DSR) dan pemodelan Unified Modeling Language (UML).
Metodologi penelitian mengikuti tahapan DSR yang mencakup identifikasi masalah, pengembangan artefak, hingga evaluasi iteratif. Sistem dirancang dengan arsitektur yang memisahkan frontend, backend, dan layanan crawler yang dikelola dalam lingkungan Docker. Fitur unggulan sistem meliputi scrapper Google Scholar yang dilengkapi algoritma eliminasi duplikasi data, kalkulasi angka kredit otomatis sesuai PO PAK, serta fitur ekspor data (Excel) untuk pelaporan. Perancangan sistem didokumentasikan secara standar menggunakan diagram UML (Use Case, Activity dengan swimlane, Sequence dengan spesifikasi objek Entity-Boundary-Control, dan Class Diagram) serta Requirement Traceability Matrix (RTM) untuk menjamin keterlacakan fitur.
Evaluasi dilakukan melalui tiga tahap: pengujian fungsional (Black Box Testing), evaluasi usability dengan klasifikasi masalah (critical, significant, nuisance), dan pengujian efisiensi waktu. Hasil pengujian menunjukkan bahwa sistem berfungsi 100% valid pada seluruh fitur utama. Evaluasi usability setelah proses iterasi menunjukkan tidak adanya masalah critical dan significant yang tersisa. Selain itu, pengujian komparatif membuktikan bahwa penggunaan sistem mampu memangkas waktu pemrosesan data DUPAK secara signifikan dibandingkan metode manual, serta menjamin akurasi perhitungan poin sesuai regulasi yang berlaku. Sistem ini menjadi solusi artefak yang efektif, efisien, dan siap diimplementasikan untuk mendukung tata kelola administrasi dosen.

=======================================

The submission and assessment process for the Credit Score Determination Proposal List (DUPAK) at the Faculty of Engineering, Universitas Muhammadiyah Surabaya (UMS) currently faces significant challenges related to time efficiency, publication data validity, and the complexity of point calculations, which must comply with the latest Operational Guidelines for Credit Score Assessment (PO PAK). The current manual method led to a high risk of data redundancy, input errors, and lengthy file verification processes. This research aimed to develop a web-based DUPAK management system that integrates web crawling techniques for publication data automation and workflow automation using n8n. This system was designed using a Design Science Research (DSR) approach and Unified Modelling Language (UML) modelling.
The research methodology followed the DSR stages, encompassing problem identification, artifact development, and iterative evaluation. The system was designed with an architecture that separates the frontend, backend, and crawler services, managed within a Docker environment. The system’s key features included a Google Scholar scraper equipped with a data duplication elimination algorithm, automatic credit score calculation according to PO PAK, and a data export feature (Excel) for reporting. The system design was documented in a standard manner using UML diagrams (Use Case, Activity with swim lanes, Sequence with Entity-Boundary-Control object specifications, and Class Diagrams) and a Requirements Traceability Matrix (RTM) to ensure feature traceability.
The evaluation was conducted in three stages: functional testing (Black Box Testing), usability evaluation with issue classification (critical, significant, and nuisance), and time efficiency testing. The test results showed that the system functioned 100% validly across all key features. Usability evaluation after the iteration process revealed no remaining critical or significant issues. Furthermore, comparative testing demonstrated that the system significantly reduced DUPAK data processing time compared to manual methods and ensured accurate point calculations in accordance with applicable regulations. This system was an effective, efficient, and ready-to-implement artifact solution to support lecturer administrative governance.