Pembentukan Model Proses Hirarki Dalam Bentuk Graph-Based Discrete Event Model Menggunakan Semantic Similarity Dan Categorization.

Ferdian, Riski Anang (2022) Pembentukan Model Proses Hirarki Dalam Bentuk Graph-Based Discrete Event Model Menggunakan Semantic Similarity Dan Categorization. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 05111840000052-Undergraduate_Thesis.pdf Restricted to Repository staff only Download (5MB)

Abstract

Model proses dapat digambarkan ke dalam dua tipe, yaitu sebuah model proses tunggal dan sebuah model proses hirarki. Model proses hirarki terdiri dari dua model, yaitu model top-level dan model bottom-level (submodel). Tujuan model proses hirarki adalah mengurangi kerumitan pada suatu model proses. Beberapa metode process discovery adalah -miner, Flexibel Heuristic Miner, Integer Linear Programming dan Fuzzy Mining. Dari keseluruhan metode process discovery yang ada, hanya sedikit metode yang mengusulkan metode pembentukan model proses hirarki dan model proses yang dapat disimulasikan. Penelitian ini mengusulkan metode process discovery baru, bernama metode Graph-based Hierarchical DES Process discovery (GHDPD), untuk membentuk sebuah model proses hirarki berbasis graf dan memetakan sebuah model proses hiraraki berbasis graf menjadi bentuk Discrete Event Simulation (DES). Tahap pertama dari metode GHDPD adalah melakukan pre-processing data untuk mendapatkan data yang bersih dari tanda baca tertentu. Kemudian, melakukan pencarian nilai similarity dari masing-masing aktivitas menggunakan metode semantic similarity. Hasil semantic similarity digunakan sebagai pedoman untuk membentuk model-model proses bottom-level dan sebuah model proses top-level. Terakhir, hasil-hasil model proses berbasis graf dipetakkan dengan aturan usulan untuk menghasilkan DES. Terdapat dua evaluasi yang dilakukan yaitu nilai kualitas model proses dan akurasi model DES terhadap model berbasis graf. Dari sisi kualitas model, model top-level metode GHDPD memiliki nilai Precision 1 dan Generalization 0,906 sedangkan model top-level Fuzzy Mining memiliki Precision 0,833 dan Generalization 0,893. Nilai kualitas model bottom-level GHDPD lebih tinggi dari model bottom-level Fuzzy Mining, dengan nilai rata-rata Fitness 1, Precision 0,952, Simplicity 1, Generalization 0,936. Secara keseluruhan, nilai kualitas model GHDPD lebih baik dari model Fuzzy Mining. Dari sisi akurasi model DES, nilai akurasi model DES versi global, top-level dan bottom-level hasil simulasi terhadap model proses berbasis graf metode GHDPD adalah 1.
=================================================================================================================================
There are two process models: a single process model and a hierarchical process model. The hierarchical process model contains a top-level model and bottom-level models (submodels). The hierarchical process model aims to reduce the complexity of a process model. Several existing process discovery methods are -miner, Flexibel Heuristic Miner, Integer Linear Programming and Fuzzy Mining. Of all those methods, only a few propose forming a hierarchical process model and a process model that can be simulated. This study offers a new process discovery method, i.e., Graph-based Hierarchical DES Process discovery (GHDPD) to form a graph-based hierarchical process model and map a graph-based hierarchical process model into a process model in the form of Discrete Event Simulation (DES). First, GHDPD pre-processes data to get clean data from certain punctuation marks. Then, GHDPD searches the similarity value of each activity using the semantic similarity method. Semantic similarity results are guidelines for forming bottom-level process models and a top-level process model. Finally, the graph-based process model is mapped with the proposed rules to produce a DES. There are two evaluations: the value of the quality of the process model and the accuracy of the DES model. Regarding model quality, the top-level model of GHDPD has a Precision of 1 and a Generalization of 0.906, while the top-level model of Fuzzy Mining has a Precision of 0.833 and a Generalization of 0.893. The quality value of the GHDPD bottom-level models is higher than the bottom-level models by Fuzzy Mining, with an average Fitness value of 1, Precision 0.952, Simplicity 1, and Generalization 0.936. Overall, the quality value of the GHDPD model is better than the Fuzzy Mining model. In terms of the accuracy of the DES model, the value of the DES model using the GHDPD method against the graph-based process model of the GHDPD method is 1.

Item Type:	Thesis (Other)
Additional Information:	RSIf 005.276 2 Fer p-1 2022
Uncontrolled Keywords:	Business Process Model, Discrete Event Simulation, Model Proses Hirarki, Process discovery. Business Process Model, Discrete Event Simulation, Hierarchical Process Model, Process discovery.
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Informatics Engineering > 55201-(S1) Undergraduate Thesis
Depositing User:	Mr. Marsudiyana -
Date Deposited:	25 May 2026 07:10
Last Modified:	25 May 2026 07:10
URI:	http://repository.its.ac.id/id/eprint/133404

Actions (login required)

View Item