Model Cross-Organizational Process Mining Untuk Mendukung Pengukuran Kinerja Procurement Korporasi

Hendrata, Ferial (2026) Model Cross-Organizational Process Mining Untuk Mendukung Pengukuran Kinerja Procurement Korporasi. Doctoral thesis, Institut Technologi Sepuluh Nopember.

Text 02411960010005-Doctoral.pdf Restricted to Repository staff only Download (22MB) | Request a copy

Abstract

Pengukuran kinerja procurement perusahaan masih banyak bertumpu pada pendekatan konvensional berbasis outcome dan KPI agregat, sehingga belum dapat menjabarkan pembentukan kinerja pada tingkat proses dan aktivitas. Sementara itu, implementasi e-procurement, khususnya e-tendering, menghasilkan event log yang merekam pelaksanaan tender secara rinci, tetapi pemanfaatannya untuk pengukuran kinerja berbasis proses, interpretasi otomatis, dan benchmarking antarentitas perusahaan dalam satu korporasi masih terbatas. Oleh karena itu, penelitian ini mengembangkan kerangka kerja cross-organizational process mining yang mengintegrasikan pengukuran kinerja berbasis event log, benchmarking antarentitas, dan interpretasi otomatis melalui agen AI berbasis large language model (LLM). Penelitian ini menggunakan pendekatan metodologis bertahap berbasis studi kasus yang mencakup perumusan masalah, persiapan data, analisis proses, dan sintesis. Data e-tendering diolah menjadi event log yang valid, konsisten, dan dapat diperbandingkan antarentitas. Process mining digunakan untuk merekonstruksi proses aktual, mengidentifikasi varian proses, serta mengukur waktu siklus dan durasi aktivitas. Hasil process mining kemudian diinterpretasikan secara otomatis melalui agen AI berbasis LLM dan digunakan sebagai dasar analitis untuk benchmarking antarentitas. Hasil penelitian menunjukkan bahwa process mining mampu menghasilkan pengukuran kinerja proses tender yang rinci, objektif, dan berbasis bukti melalui identifikasi alur proses aktual, varian proses, waktu siklus, durasi aktivitas, serta identifikasi indikasi bottleneck proses. Agen AI berbasis LLM mampu mentransformasikan model dan statistik kinerja proses menjadi narasi analitis yang sistematis dan mudah dipahami secara manajerial. Integrasi keduanya menghasilkan kerangka kerja cross-organizational process mining untuk membandingkan kinerja proses tender antarentitas, mengidentifikasi kesenjangan kinerja, dan mendukung rekomendasi perbaikan berbasis data. Kontribusi utama penelitian ini terletak pada penyatuan pengukuran kinerja berbasis event log, interpretasi otomatis berbasis LLM, dan benchmarking antarentitas dalam satu kerangka analitis untuk mendukung pengambilan keputusan manajerial pada procurement korporasi.
=================================================================================================================================
The measurement of procurement performance in companies still heavily relies on conventional approaches based on outcomes and aggregate KPIs, thus failing to delineate performance formation at the process and activity levels. Meanwhile, the implementation of e-procurement, particularly e-tendering, generates event logs that record tender execution in detail, but its utilization for process-based performance measurement, automatic interpretation, and benchmarking between entities within a single corporation remains limited. Therefore, this research develops a cross-organizational process mining framework that integrates performance measurement based on event logs, inter-entity benchmarking, and automatic interpretation through AI agents based on large language models (LLM). This research uses a phased methodological approach based on case studies that includes problem formulation, data preparation, process analysis, and synthesis. E-tendering data is processed into a valid, consistent, and comparable event log between entities. Process mining is used to reconstruct actual processes, identify process variants, and measure cycle times and activity durations. The results of process mining are then automatically interpreted through an LLM-based AI agent and used as an analytical basis for benchmarking between entities. The research results show that process mining is capable of producing detailed, objective, and evidence-based measurements of tender process performance through the identification of actual process flows, process variants, cycle times, activity durations, and identification of process bottleneck indications. The LLM-based AI agent can transform process performance models and statistics into systematic and easily understandable managerial analytical narratives. The integration of both results in a cross-organizational process mining framework to compare tender process performance between entities, identify performance gaps, and support data-driven improvement recommendations. The main contribution of this research lies in the integration of event log-based performance measurement, LLM-based automatic interpretation, and inter-entity benchmarking into a single analytical framework to support managerial decision-making in corporate procurement.

Item Type:	Thesis (Doctoral)
Uncontrolled Keywords:	cross-organizational process mining, e-tendering, pengukuran kinerja procurement, agen AI berbasis LLM, benchmarking ================= Keywords: cross-organizational process mining, e-tendering, procurement performance measurement, LLM-based AI agents, benchmarking
Subjects:	T Technology > T Technology (General) > T57.5 Data Processing T Technology > T Technology (General) > T58.6 Management information systems
Divisions:	Faculty of Industrial Technology and Systems Engineering (INDSYS) > Industrial Engineering > 26001-(S3) PhD Thesis
Depositing User:	Ferial Hendrata
Date Deposited:	24 Jul 2026 06:08
Last Modified:	24 Jul 2026 06:08
URI:	http://repository.its.ac.id/id/eprint/137135

Actions (login required)

View Item