Sarwono, Armadya Hermawan (2025) Metode Process Discovery Berbasis Graf dengan Pendekatan Heuristic Miner untuk Menyederhanakan Visualisasi Model Proses. Other thesis, Institut Teknologi Sepuluh Nopember.
![[thumbnail of 5025211243-Undergraduate_Thesis.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
5025211243-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (7MB) | Request a copy |
Abstract
Process mining merupakan teknik analisis yang penting untuk memahami dan mengoptimalkan proses bisnis berdasarkan data event log. Penelitian ini mengusulkan metode process discovery berbasis graf dengan pendekatan heuristic miner untuk menyederhanakan visualisasi model proses. Metode graph-based heuristic miner dirancang untuk mengatasi keterbatasan metode konvensional dalam menghasilkan model yang kompleks dan sulit diinterpretasi. Pendekatan ini memanfaatkan platform Neo4j untuk menyimpan dan memvisualisasikan model proses dalam bentuk graf, sehingga memungkinkan representasi hubungan antar aktivitas yang lebih jelas dan mudah dipahami. Evaluasi penelitian ini menggunakan tiga dataset, yaitu Simple Event Log yang memiliki 14 traces, Solid Medical Waste Handling yang memiliki 11 traces, dan BPI Challenge 2019 yang memiliki 1499 traces. Hasil evaluasi menggunakan metode graph-based heuristic miner menunjukkan performa yang bervariasi berdasarkan karakteristik dataset. Pada dataset Simple Event Log, metode ini menghasilkan nilai fitness 1,00, precision 0,218, generalization 0,891, dan simplicity 1,00 dengan dependency threshold. Dataset Solid Medical Waste Handling menghasilkan nilai fitness 0,909, precision 0,550, generalization 0,804, dan simplicity 1,00. Sementara pada dataset BPI Challenge 2019, metode menghasilkan simplicity 1,00. Penelitian ini berhasil mengembangkan metode yang menghasilkan visualisasi model proses yang lebih sederhana dan informatif dibandingkan metode heuristic miner konvensional.
===================================================================================================================================
Process mining is an important analytical technique for understanding and optimizing business processes based on event log data. This research proposes a graph-based process discovery method with a heuristic miner approach to simplify process model visualization. The graph-based heuristic miner method is designed to overcome the limitations of conventional methods in producing complex and difficult-to-interpret models. This approach utilizes the Neo4j platform to store and visualize process models in graph form, enabling clearer and more understandable representation of relationships between activities. The evaluation of this research uses three datasets: Simple Event Log with 14 traces, Solid Medical Waste Handling with 11 traces, and BPI Challenge 2019 with 1499 traces. The evaluation results using the graph-based heuristic miner method show varying performance based on dataset characteristics. On the Simple Event Log dataset, this method produces fitness value of 1.00, precision 0.218, generalization 0.891, and simplicity 1.00 with dependency threshold. The Solid Medical Waste Handling dataset produces fitness value of 0.909, precision 0.550, generalization 0.804, and simplicity 1.00. Meanwhile, on the BPI Challenge 2019 dataset, the method produces simplicity 1.00. This research successfully develops a method that produces simpler and more informative process model visualization compared to conventional heuristic miner methods.
| Item Type: | Thesis (Other) |
|---|---|
| Uncontrolled Keywords: | Business Process Model, Graph-based Mining, Heuristic Miner, Neo4j, Process Discovery Graph-based Mining, Heuristic Miner, Model Proses Bisnis, Process Discovery |
| Subjects: | Q Science > QA Mathematics > QA166 Graph theory Q Science > QA Mathematics > QA76.9.I52 Information visualization |
| Divisions: | Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Informatics Engineering > 55201-(S1) Undergraduate Thesis |
| Depositing User: | Armadya Hermawan Sarwono |
| Date Deposited: | 31 Jul 2025 07:40 |
| Last Modified: | 31 Jul 2025 07:40 |
| URI: | http://repository.its.ac.id/id/eprint/124178 |
Actions (login required)
 |
View Item |