Raphael Zefferer
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Raphael Zefferer has a total of 5 co-authors. Here are the top co-authors by number of shared publications:
- Bernhard Schenkenfelder
Software Competence Center Hagenberg GmbH, Hagenberg, Austria
4 shared publications - Stefan Wagner
University of Applied Sciences Upper Austria, Hagenberg, Austria
2 shared publications - Christian Salomon
Software Competence Center Hagenberg GmbH
1 shared publication - Manuel Wimmer
Johannes Kepler University Linz, Linz, Austria
1 shared publication - Mario Winterer
Software Competence Center Hagenberg GmbH
1 shared publication
Latest Publications
- Open Access
- 4 October 2026
- Published by Association for Computing Machinery (ACM)
- p. 289-293
Code2LowCode: An Interactive Tool for Lifting Code Clones into Reusable Abstractions
Code duplication is widespread in software, yet the dominant response, clone detection, stops at flagging duplication and leaves the harder, value-creating step to developers: turning recurring code into reusable abstractions. We argue that recent advances in Large Language Models (LLMs), combined with classical clone analysis and the rich abstraction vocabulary of Model-Driven Engineering (MDE), make it possible to begin closing this gap automatically. We present Code2LowCode, a prototype that connects clone detection to MDE by lifting detected clones into reusable, validated artifacts. Code2LowCode combines two ideas. First, a cascaded hybrid detection pipeline routes candidate clone pairs through increasingly expensive tiers, from classical detectors to code embeddings to LLM classifiers. This ensures that cheap tiers resolve clear cases (Types 1-3), while only ambiguous, potentially semantic (Type 4) pairs reach costly analysis. Second, an LLM-driven, human-in-the-loop workspace analyzes the resulting clone set and proposes a unified abstraction strategy, emitting concrete MDE artifacts, such as parameterized templates, ANTLR4 DSL grammars, EMF Ecore metamodels, and JSON schemas. Each artifact is validated (grammar compilation, javac, schema checks) and approved by a developer before target code is generated in an agentic workflow. This reframes clone management not as defect-hunting but as a path toward automated low-code asset creation.
- 4 October 2026
- Published by Association for Computing Machinery (ACM)
- p. 289-293
Open AccessCode2LowCode: An Interactive Tool for Lifting Code Clones into Reusable Abstractions
Code duplication is widespread in software, yet the dominant response, clone detection, stops at flagging duplication and leaves the harder, value-creating step to developers: turning recurring code into reusable abstractions. We argue that recent advances in Large Language Models (LLMs), combined with classical clone analysis and the rich abstraction vocabulary of Model-Driven Engineering (MDE), make it possible to begin closing this gap automatically. We present Code2LowCode, a prototype that connects clone detection to MDE by lifting detected clones into reusable, validated artifacts. Code2LowCode combines two ideas. First, a cascaded hybrid detection pipeline routes candidate clone pairs through increasingly expensive tiers, from classical detectors to code embeddings to LLM classifiers. This ensures that cheap tiers resolve clear cases (Types 1-3), while only ambiguous, potentially semantic (Type 4) pairs reach costly analysis. Second, an LLM-driven, human-in-the-loop workspace analyzes the resulting clone set and proposes a unified abstraction strategy, emitting concrete MDE artifacts, such as parameterized templates, ANTLR4 DSL grammars, EMF Ecore metamodels, and JSON schemas. Each artifact is validated (grammar compilation, javac, schema checks) and approved by a developer before target code is generated in an agentic workflow. This reframes clone management not as defect-hunting but as a path toward automated low-code asset creation.
- Open Access
- 4 October 2026
- Published by Association for Computing Machinery (ACM)
- p. 732-736
From Code to Low-Code: An LLM-Driven Pipeline for Lifting Code Clones into Reusable Abstractions
Low-Code Development Platforms (LCDPs) promise substantial efficiency gains by allowing citizen developers to create applications from reusable building blocks instead of manually writing code. Realizing these benefits, however, depends on having such abstractions in the first place. Therefore, a curated library of reusable components is required, building this library remains a persistent challenge though. Recurring logic is often scattered as duplicated code across different systems in the same domain. Turning it into reusable assets requires language engineers to inspect large amounts of code, distinguish incidental from meaningful similarity, choose an appropriate abstraction mechanism, and implement it correctly. This effort limits the adoption of low-code approaches. To address this, we draw on principles from Low-Code Development (LCD) and Model-Driven Engineering (MDE) and introduce an LLM-driven approach for transforming code clones into reusable abstractions. Our approach combines three stages: (i) a hybrid clone detection pipeline that pairs classical detectors with code embeddings and LLM classifiers; (ii) a human-in-the-loop abstraction stage that proposes a unified abstraction strategy; and (iii) an agentic generation stage that synthesizes and validates concrete artifacts such as parameterized templates, DSL grammars, entity models, and schemas. These artifacts preserve the semantics of the original code while reducing complexity through guided, conversational refinement and automated validation before adoption. This paper is scoped to the lifting itself, that is, stages (ii) and (iii), and treats detection as the upstream supplier of their input. We evaluate these two stages on a codebase drawn from a real-world industry case, assessing the feasibility of the workflow, the division of labor between engineer and model, and the fidelity and reusability of the synthesized abstractions, and report first evidence that LLM-driven lifting can produce adoption-ready artifacts with reduced manual effort.
- 4 October 2026
- Published by Association for Computing Machinery (ACM)
- p. 732-736
Open AccessFrom Code to Low-Code: An LLM-Driven Pipeline for Lifting Code Clones into Reusable Abstractions
Low-Code Development Platforms (LCDPs) promise substantial efficiency gains by allowing citizen developers to create applications from reusable building blocks instead of manually writing code. Realizing these benefits, however, depends on having such abstractions in the first place. Therefore, a curated library of reusable components is required, building this library remains a persistent challenge though. Recurring logic is often scattered as duplicated code across different systems in the same domain. Turning it into reusable assets requires language engineers to inspect large amounts of code, distinguish incidental from meaningful similarity, choose an appropriate abstraction mechanism, and implement it correctly. This effort limits the adoption of low-code approaches. To address this, we draw on principles from Low-Code Development (LCD) and Model-Driven Engineering (MDE) and introduce an LLM-driven approach for transforming code clones into reusable abstractions. Our approach combines three stages: (i) a hybrid clone detection pipeline that pairs classical detectors with code embeddings and LLM classifiers; (ii) a human-in-the-loop abstraction stage that proposes a unified abstraction strategy; and (iii) an agentic generation stage that synthesizes and validates concrete artifacts such as parameterized templates, DSL grammars, entity models, and schemas. These artifacts preserve the semantics of the original code while reducing complexity through guided, conversational refinement and automated validation before adoption. This paper is scoped to the lifting itself, that is, stages (ii) and (iii), and treats detection as the upstream supplier of their input. We evaluate these two stages on a codebase drawn from a real-world industry case, assessing the feasibility of the workflow, the division of labor between engineer and model, and the fidelity and reusability of the synthesized abstractions, and report first evidence that LLM-driven lifting can produce adoption-ready artifacts with reduced manual effort.
- 20 June 2026
- Published by Springer Nature
- p. 329-332
A Web-Based Demonstrator for Quantum Circuit Programming Education
- 20 June 2026
- Published by Springer Nature
- p. 329-332
A Web-Based Demonstrator for Quantum Circuit Programming Education
- 13 June 2025
- Published by Springer Nature
- p. 99-107
Assembly Workers as PLC Programmers: What End Users Need to Understand About Low-Code Development Platforms
- 13 June 2025
- Published by Springer Nature
- p. 99-107
Assembly Workers as PLC Programmers: What End Users Need to Understand About Low-Code Development Platforms