MEASURING BUSINESS PROCESS SIMILARITY USING PROBABILISTIC LATENT SEMANTIC ANALYSIS (PLSA) AND GREEDY GRAPH MATCHING

Fildzah Shabrina, Riyanarto Sarno

Abstract


Abstract
The business process is a set of activities and tasks performed to achieve the goals of an organization. The business process model can be reused as a business process management effort into a repository. To solve the problem, it is necessary to measure the business process model that has similarity or similarity in terms of activity or process. From several business process models that have similarity can be identified as the main business process model, which has the primary function of the same activity. Business process model matching is the one of technique that can be used to identify, to measure the similarity of a set of business process models. The graph matching approach fit to identify the similarity of processes or activities in the business process model. The technique of matching the graph with Greedy graph matching shows similar results with an 89% precision value based on measuring the similarity of the graph building structure. Another approach in graph matching is a semantically or a text-based. Probabilistic Latent Semantic Analysis (PLSA) is one of the semantic approaches to measure the similarity of text in documents. PLSA measures the linkage of words in the document to identify any similarity of topics in the document. Measuring PLSA in business process matching analysis is by comparing text labels on each node in the business process. This research measures the similarity of business process models by combining two similarity analysis techniques based on semantics using PLSA and structural with Greedy. A graph matching technique by computing the semantics of each label on activities that are related to other activity labels. Structurally, connected activities are related to the same process or the same function. The result of this research is to know the effectiveness of business process which has activity relation.
Keywords : Business Process, BPMN, Graph Similarity, Probabilistic Latent Semantic Analysis (PLSA), Greedy Graph Matching


Proses bisnis adalah serangkaian aktivitas dan tugas yang dilakukan untuk mencapai tujuan dari sebuah organisasi. Model proses bisnis dapat digunakan kembali sebagai upaya manajemen proses bisnis tersebut ke dalam sebuah repositori. Dalam repositori berisi ratusan hingga ribuan model proses bisnis dengan model yang sama maupun berbeda. Hingga dapat terjadinya duplikasi dan penumpukkan data. Untuk mengatasi permasalahan tersebut, perlunya dilakukan pengukuran terhadap model proses bisnis yang memiliki kesamaan atau kemiripan dalam hal aktivitas ataupun proses. Beberapa model proses bisnis yang memiliki kemiripan (similarity) dapat diidentifikasi sebagai model proses bisnis utama, yaitu memiliki fungsi dan aktivitas yang sama. Mencocokkan model proses bisnis merupakan salah satu teknik untuk mengidentifikasi, mengukur kemiripan dari kumpulan model proses bisnis. Pendekatan pencocokkan graf (graph matching) cocok untuk mengidentifikasi kemiripan proses atau aktivitas dalam model proses bisnis. Teknik mencocokkan graf dengan Greedy graph matching menghasilkan nilai presisi sebesar 89% berdasarkan pengukuran kemiripan struktur graf. Pendekatan lain dalam pencocokkan graf ialah secara semantik atau teks. Probabilistic Latent Semantic Analysis (PLSA) merupakan salah satu pendekatan semantik untuk menghitung kemiripan teks dalam dokumen. Perhitungan PLSA dalam analisis pencocokkan proses bisnis adalah dengan membandingkan label teks pada tiap node (label) proses bisnis. Penelitian ini mengukur kemiripan model proses bisnis dengan menggabungkan dua teknik analisis kemiripan berdasarkan semantik menggunakan PLSA dan struktural dengan Greedy. Teknik pencocokkan graf dengan menghitung semantik dari setiap label aktivitas yang saling memiliki keterkaitan atau hubungan. Secara struktural, beberapa aktivitas saling terhubung memiliki keterkaitan proses atau fungsi yang sama. Hasil penelitian ini adalah untuk mengetahui efektifitas dari proses bisnis yang memiliki keterkaitan aktivitas.
Kata Kunci : Proses Bisnis, BPMN, Kemiripan Graf, Probabilistic Latent Semantic Analysis (PLSA), Greedy Graph Matching.


Keywords


Business Process; BPMN; Graph Similarity; Probabilistic Latent Semantic Analysis (PLSA); Greedy Graph Matching

Full Text:

PDF

References


Anugrah, I.G. & Sarno, R. 2016. Business Process Model Similarity Analysis Using Hybrid PLSA and WDAG Methods, pp. 231–236.

Becker, M. & Laue, R. 2011. “Analysing Differences Between Business Process Similarity Measures”, in International Conference on Business Process.

Djikman, R., Dumas, M., Dongen, B.V., Reina, K., & Mendling, J. 2011. Similarity of business process models : Metrics and evaluation, vol. 36, pp. 498–516.

Djikman, R., Dumas, M., & Garc, L. 2009. “Graph Matching Algorithms for Business Process Model Similarity Search,” in International Conference on Business Process Management, pp. 48–63.

Djikman, R., & Dumas, M. 2009. Aligning Business Process Models ¨,” in Enterprise Distributed Object Computing Conference, pp. 1–16.

Hofmann, T. 1999. “Probabilistic Latent Semantic Analysis,” Proc. Fifteenth Conf.

Klinkmüller, C. & Weber, I. 2017. Analyzing control flow information to improve the effectiveness of process model matching techniques, Decis. Support Syst., vol. 100, pp. 6– 14. .

Kuss, E. & Stuckenschmidt, H. 2017. “Automatic Classification to Matching Patterns for Process Model Matching Evaluation,” in CEUR Workshop Proceedings 1979, pp. 306–319.

Weidlich, M., Djikman, R., & Mendling, J. “The ICoP Framework : Identification of Correspondences between Process Models”.


Refbacks

  • There are currently no refbacks.
slot gacor slot gacor hari ini slot gacor 2025 demo slot pg slot gacor slot gacor