Metode Graph-based Process Merging untuk Menggabungkan Proses Bisnis Mengandung Invisible Task

Ahmad, Syamil Hibban (2025) Metode Graph-based Process Merging untuk Menggabungkan Proses Bisnis Mengandung Invisible Task. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 05111840000041-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (3MB) | Request a copy

Abstract

Di bidang manajemen proses bisnis, penerapan strategi membangun model proses yang efisien dapat meningkatkan kualitas proses bisnis perusahaan. Penggunaan kembali model proses bisnis yang sudah ada atau menggabungkan dua model proses bisnis merupakan pendekatan praktis untuk membangun model proses bisnis. Salah satu metode process mining yang ada yaitu Graph-based Process Merging dapat menggabungkan dua model proses bisnis. Kekurangan dari metode Graph-based Process Merging adalah belum dapat melakukan penggabungan model proses bisnis yang mengandung invisible task. Tugas akhir ini mengusulkan pengembangan metode Graph-based Process Merging untuk membentuk rule yang menggabungkan model proses yang memiliki relasi invisible task maupun model proses yang hasil penggabungannya menghasilkan invisible task. Invisible task merupakan aktivitas yang tidak muncul di event log tetapi ditambahkan dalam model proses untuk menggambarkan kondisi tertentu. Invisible task terdiri dari invisible prime task untuk menggambarkan kondisi proses skip dan redo, sedangkan invisible non-prime task untuk menggambarkan kondisi penumpukan relasi paralel. Evaluasi dilakukan dengan membandingkan hasil model proses gabungan dari metode yang diusulkan dengan model proses gabungan dari Graph-based Process Merging dan metode Derguech menggunakan pengukuran akurasi. Berdasarkan akurasi hasil model proses gabungan, hasil model dari metode usulan memiliki rata-rata akurasi 1,00, sedangkan akurasi rata-rata dari metode Graph-based Process Merging adalah 0,4 dan metode Derguech memiliki akurasi rata-rata 0,83.
====================================================================================================================================
In the field of business process management, implementing a strategy to build efficient process models can enhance the business processes quality of a company. Reusing existing business process models or combining two business process models is a practical approach to building business process models. One existing process mining method, Graph-based Process Merging, can combine two business process models. However, the Graph-based Process Merging method is limited in its ability to merge models containing invisible tasks. This thesis proposes the development of a Graph-based Process Merging method to form rules for merging process models that have invisible task relationships or process models whose merged results produce invisible tasks. Invisible tasks are activities that do not appear in the event log but are added to the process model to describe certain conditions. Invisible tasks consist of invisible prime tasks to describe process skip and redo conditions, while invisible non-prime tasks describe parallel relationship accumulation conditions. The evaluation was conducted by comparing the resulting merged process model from the proposed method with the merged process models from Graph-based Process Merging and the Derguech method using accuracy measurements. Based on the accuracy of the merged process models, the results from the proposed method have an average accuracy of 1.00, while the average accuracy of the Graph-based Process Merging method is 0.4 and the Derguech method has an average accuracy of 0.83.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Graph-based method, Invisible task, Merging business process, Model proses, Graph-based method, Invisible task, Merging business process, Process model
Subjects:	T Technology > T Technology (General) > T57.5 Data Processing
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Informatics Engineering > 55201-(S1) Undergraduate Thesis
Depositing User:	Syamil Hibban Ahmad
Date Deposited:	24 Jul 2025 03:41
Last Modified:	24 Jul 2025 03:41
URI:	http://repository.its.ac.id/id/eprint/120948

Actions (login required)

View Item