Pemanfaatan Large Language Model untuk Pembangkitan Model Domain Menggunakan Pendekatan Tree-of-Thoughts

Putri, Helsa Sriprameswari (2026) Pemanfaatan Large Language Model untuk Pembangkitan Model Domain Menggunakan Pendekatan Tree-of-Thoughts. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5025221154-Undergraduate_Thesis.pdf Restricted to Repository staff only Download (4MB)

Abstract

Pemodelan domain merupakan tahap penting Domain-Driven Design (DDD) dalam menghasilkan rich domain model yang merepresentasikan aturan bisnis. Pendekatan otomatis untuk membangkitkan model domain telah banyak dikembangkan, tetapi masih menghasilkan anemic domain model yang belum mencakup elemen Tactical DDD. Oleh karena itu, diperlukan pendekatan otomatis untuk membangkitkan rich domain model, sehingga proses pemodelan domain menjadi lebih efisien. Tugas akhir ini mengusulkan pendekatan otomatis untuk menghasilkan rich domain model menggunakan Large Language Model (LLM) dengan metode Tree-of-Thoughts (ToT). Metode ToT mampu melakukan reasoning bertahap melalui eksplorasi solusi untuk menghasilkan elemen domain meliputi Business Rules, Value Object, Entity, Aggregate dan Aggregate Root, Repository, Domain Service, serta Domain Event. Evaluasi dilakukan pada delapan dataset domain requirement, yaitu BTMS, H2S, LabTracker, CelO, TSS, SHAS, OTS, dan HBMS, dengan membandingkan model OpenAI GPT-4 dan GPT-5 menggunakan variasi temperatur serta lima kali percobaan berulang. Kinerja model dievaluasi menggunakan metrik precision, recall, dan F1-score. Hasil menunjukkan bahwa GPT-4 dengan temperatur 0,5 menghasilkan performa terbaik dengan rata-rata F1-score sebesar 0,88. Sementara itu, GPT-5 cenderung memiliki recall lebih tinggi, tetapi diikuti penurunan precision. Hasil penelitian menunjukkan bahwa kombinasi LLM dan ToT mampu menghasilkan rich domain model secara otomatis dengan tingkat akurasi yang baik.
=====================================================================================================================================
Domain modeling is a fundamental phase in Domain-Driven Design (DDD) for developing a rich domain model that represents the business rules of a software system. Automatic approaches for domain model generation have been widely developed. However, they still tend to produce anemic domain models that lack Tactical DDD elements. Therefore, an automated approach is needed to improve the efficiency of the domain modeling process while producing more complete rich domain models. This undergraduate thesis proposes an automated approach for generating rich domain models using a Large Language Model (LLM) with a Tree-of-Thoughts (ToT) approach. The ToT method performs reasoning through the exploration of solution paths to generate domain model elements, including business rules, value object, entity, aggregate and aggregate root, repository, domain service, and domain event. The proposed approach was evaluated using eight domain requirement datasets, namely BTMS, H2S, LabTracker, CelO, TSS, SHAS, OTS, and HBMS. The evaluation compared OpenAI GPT-4 and GPT-5 under different temperature settings and across five repeated experimental runs. Model performance was assessed using precision, recall, and F1-score. The results indicate that GPT-4 with a temperature of 0.5 achieved the best performance, obtaining an average F1-score of 0.88. In contrast, GPT-5 generally achieved higher recall but at the expense of lower precision. Overall, the findings demonstrate that the combination of LLMs and the ToT method can automatically generate rich domain models with a high level of accuracy.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Domain-Driven Design, Large Language Model, Otomatisasi Pemodelan Perangkat Lunak, Pembangkitan Domain Model, Tree-of-Thoughts
Subjects:	T Technology > T Technology (General)
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Informatics Engineering > 55201-(S1) Undergraduate Thesis
Depositing User:	Helsa Sriprameswari Putri
Date Deposited:	29 Jul 2026 02:31
Last Modified:	29 Jul 2026 02:31
URI:	http://repository.its.ac.id/id/eprint/139269

Available Versions of this Item

• Pemanfaatan Large Language Model untuk Pembangkitan Model Domain Menggunakan Pendekatan Tree-of-Thoughts. (deposited 28 Jul 2026 02:25)
  • Pemanfaatan Large Language Model untuk Pembangkitan Model Domain Menggunakan Pendekatan Tree-of-Thoughts. (deposited 29 Jul 2026 02:31) [Currently Displayed]

Actions (login required)

View Item