STAF 2025 (series) / OCL 2025 (series) / OCL 2025 /
Using LLMs to Extract OCL Specifications from Java and Python Programs: An Empirical Study
This program is tentative and subject to change.
Tue 10 Jun 2025 11:40 - 12:05 at D 239 - OCL Session 1: OCL and beyond
This paper presents a comprehensive study of the application of several open-source Large Language Models (LLMs) for extracting Object Constraint Language (OCL) specifications from source code. We aim to provide researchers and developers with insights into the capabilities and limitations of using different LLMs to extract OCL constraints from code.
We evaluate a collection of open-source LLMs of comparable size (StarCoder2, LLaMA, CodeLlama, Mistral, and DeepSeek) by prompting them to generate OCL constraints for both Java and Python programs. The results show that both Mistral and DeepSeek outperform other LLMs in extracting OCL constraints from both languages.
This program is tentative and subject to change.
Tue 10 JunDisplayed time zone: Amsterdam, Berlin, Bern, Rome, Stockholm, Vienna change
Tue 10 Jun
Displayed time zone: Amsterdam, Berlin, Bern, Rome, Stockholm, Vienna change
11:00 - 12:30 | |||
11:00 5mDay opening | Welcome OCL | ||
11:05 10mTalk | OCL on Life Support: Can We Revitalize the Community for a Stronger Future? OCL Mauro Dalle Lucca Tosi Luxembourg Institute of Sciences and Technology, Jordi Cabot Luxembourg Institute of Science and Technology | ||
11:15 25mTalk | Sorted collection and map types for OCL OCL Dr Kevin Lano King's College London, Shekoufeh Rahimi University of Roehampton , Zishan Rahman King's College London | ||
11:40 25mTalk | Using LLMs to Extract OCL Specifications from Java and Python Programs: An Empirical Study OCL | ||
12:05 25mTalk | From OCL to JSX: Declarative Constraint Modeling in Modern SaaS Tools OCL Antonio Bucchiarone DISIM, University of L'Aquila, Juri Di Rocco University of L'Aquila, Damiano Di Vincenzo Università degli Studi dell'Aquila, Alfonso Pierantonio | ||