Sebastiano Antonio Piccolo
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Sebastiano Antonio Piccolo has a total of 50 co-authors. Here are the top co-authors by number of shared publications:
- Anja Maier
University of Strathclyde
5 shared publications - Antonino Quattrocchi
Department of Engineering, University of Messina, 98166 Messina, Italy
3 shared publications - Pasquale De Meo
University of Messina
3 shared publications - Sune Lehmann
Department of Applied Mathematics and Computer Science, Technical University of Denmark
3 shared publications - Emilio Ferrara
University of Southern California, Los Angeles, CA, USA
2 shared publications - Giorgio Terracina
University of Calabria
2 shared publications
Latest Publications
- Open Access
- 28 September 2026
- Published by AI Access Foundation in Journal of Artificial Intelligence Research
- Vol. 87
Integrating Answer Set Programming and Large Language Models for Reliable Structured Knowledge Extraction
Answer Set Programming (ASP) and Large Language Models (LLMs) offer complementary strengths in symbolic reasoning and natural language understanding. In this work, we present an integrated framework for reliable structured knowledge extraction from natural language that combines the syntactic capabilities of LLMs with the semantic constraints provided by ASP. Structured representations in the form of ASP facts are obtained by guiding LLM generation through domain specific prompt templates and background knowledge specified in a declarative YAML configuration, and by validating and completing the extracted information using an ASP knowledge base. Building on this framework, we introduce grammar constrained decoding as a principled mechanism to enforce structural and semantic alignment between LLM outputs and target symbolic representations. By leveraging formal grammars, we gain fine grained control over the generation process, substantially reducing hallucinations and limiting unnecessary verbosity, while preserving extraction quality. We study three grammar formats commonly used to represent structured knowledge, namely CSV, Datalog, and JSON, and analyze their impact on correctness, efficiency, and computational cost. We evaluate the proposed approach on benchmarks derived from ASP Competitions. The results show that the integration of ASP consistently improves extraction accuracy over vanilla LLMs, particularly for smaller models. Moreover, grammar constrained decoding maintains these gains while significantly improving generation efficiency. Among the considered formats, CSV provides the best trade-off between accuracy and cost, JSON achieves the highest extraction quality at the expense of increased verbosity, and Datalog exhibits lower robustness in the extraction process. Overall, the results demonstrate that combining ASP with grammar constrained LLM output yields a reliable, scalable, and cost-effective approach to structured knowledge extraction.
- 28 September 2026
- Published by AI Access Foundation in Journal of Artificial Intelligence Research
- Vol. 87
Open AccessIntegrating Answer Set Programming and Large Language Models for Reliable Structured Knowledge Extraction
Answer Set Programming (ASP) and Large Language Models (LLMs) offer complementary strengths in symbolic reasoning and natural language understanding. In this work, we present an integrated framework for reliable structured knowledge extraction from natural language that combines the syntactic capabilities of LLMs with the semantic constraints provided by ASP. Structured representations in the form of ASP facts are obtained by guiding LLM generation through domain specific prompt templates and background knowledge specified in a declarative YAML configuration, and by validating and completing the extracted information using an ASP knowledge base. Building on this framework, we introduce grammar constrained decoding as a principled mechanism to enforce structural and semantic alignment between LLM outputs and target symbolic representations. By leveraging formal grammars, we gain fine grained control over the generation process, substantially reducing hallucinations and limiting unnecessary verbosity, while preserving extraction quality. We study three grammar formats commonly used to represent structured knowledge, namely CSV, Datalog, and JSON, and analyze their impact on correctness, efficiency, and computational cost. We evaluate the proposed approach on benchmarks derived from ASP Competitions. The results show that the integration of ASP consistently improves extraction accuracy over vanilla LLMs, particularly for smaller models. Moreover, grammar constrained decoding maintains these gains while significantly improving generation efficiency. Among the considered formats, CSV provides the best trade-off between accuracy and cost, JSON achieves the highest extraction quality at the expense of increased verbosity, and Datalog exhibits lower robustness in the extraction process. Overall, the results demonstrate that combining ASP with grammar constrained LLM output yields a reliable, scalable, and cost-effective approach to structured knowledge extraction.
- Open Access
- 6 March 2026
- Published by Elsevier in Information Sciences
- Vol. 743
The theory and practice of computing the bus factor
- 6 March 2026
- Published by Elsevier in Information Sciences
- Vol. 743
Open AccessThe theory and practice of computing the bus factor
- Open Access
Fast and Accurate Heuristics for Bus-Factor Estimation
The bus-factor is a critical risk indicator that quantifies how many key contributors a project can afford to lose before core knowledge or functionality is compromised. Despite its practical importance, accurately computing the bus-factor is NP-Hard under established formalizations, making scalable analysis infeasible for large software systems. In this paper, we model software projects as bipartite graphs of developers and tasks and propose two novel approximation heuristics, Minimum Coverage and Maximum Coverage, based on iterative graph peeling, for two influential bus-factor formalizations. Our methods significantly outperform the widely adopted degree-based heuristic, which we show can yield severely inflated estimates. We conduct a comprehensive empirical evaluation on over $1\,000$ synthetic power-law graphs and demonstrate that our heuristics provide tighter estimates while scaling to graphs with millions of nodes and edges in minutes. Our results reveal that the proposed heuristics are not only more accurate but also robust to structural variations in developer-task assignment graph. We release our implementation as open-source software to support future research and practical adoption.Open AccessFast and Accurate Heuristics for Bus-Factor Estimation
The bus-factor is a critical risk indicator that quantifies how many key contributors a project can afford to lose before core knowledge or functionality is compromised. Despite its practical importance, accurately computing the bus-factor is NP-Hard under established formalizations, making scalable analysis infeasible for large software systems. In this paper, we model software projects as bipartite graphs of developers and tasks and propose two novel approximation heuristics, Minimum Coverage and Maximum Coverage, based on iterative graph peeling, for two influential bus-factor formalizations. Our methods significantly outperform the widely adopted degree-based heuristic, which we show can yield severely inflated estimates. We conduct a comprehensive empirical evaluation on over $1\,000$ synthetic power-law graphs and demonstrate that our heuristics provide tighter estimates while scaling to graphs with millions of nodes and edges in minutes. Our results reveal that the proposed heuristics are not only more accurate but also robust to structural variations in developer-task assignment graph. We release our implementation as open-source software to support future research and practical adoption. - 24 January 2025
- Published by Springer Nature
- p. 255-270
Evaluating and Improving Projects’ Bus-Factor: A Network Analytical Framework
- 24 January 2025
- Published by Springer Nature
- p. 255-270
Evaluating and Improving Projects’ Bus-Factor: A Network Analytical Framework
- Open Access
- 19 June 2023
- Published by Association for Computing Machinery (ACM)
- p. 217-226
On predicting and explaining asylum adjudication
- 19 June 2023
- Published by Association for Computing Machinery (ACM)
- p. 217-226
Open AccessOn predicting and explaining asylum adjudication
- 7 February 2023
- Published by Oxford University Press (OUP) in Journal of Refugee Studies
- Vol. 37 (4) , 915-937
Data-driven futures of international refugee law
As refugee law practice enters the world of data, it is time to take stock as to what refugee law research can gain from technological developments. This article provides an outline for a computationally driven research agenda to tackle refugee status determination variations as a recalcitrant puzzle of refugee law. It first outlines how the growing field of computational law may be canvassed to conduct legal research in refugee studies at a greater empirical scale than traditional legal methods. It then turns to exemplify the empirical purchase of a data-driven approach to refugee law through an analysis of the Danish Refugee Appeal Board’s asylum case law and outlines methods for comparison with datasets from Australia, Canada, and the United States. The article concludes by addressing the data politics arising from a turn to digital methods, and how these can be confronted through insights from critical data studies and reflexive research practices.
- 7 February 2023
- Published by Oxford University Press (OUP) in Journal of Refugee Studies
- Vol. 37 (4) , 915-937
Data-driven futures of international refugee law
As refugee law practice enters the world of data, it is time to take stock as to what refugee law research can gain from technological developments. This article provides an outline for a computationally driven research agenda to tackle refugee status determination variations as a recalcitrant puzzle of refugee law. It first outlines how the growing field of computational law may be canvassed to conduct legal research in refugee studies at a greater empirical scale than traditional legal methods. It then turns to exemplify the empirical purchase of a data-driven approach to refugee law through an analysis of the Danish Refugee Appeal Board’s asylum case law and outlines methods for comparison with datasets from Australia, Canada, and the United States. The article concludes by addressing the data politics arising from a turn to digital methods, and how these can be confronted through insights from critical data studies and reflexive research practices.
- Open Access
- 5 December 2022
- Published by IOS Press
- p. 164-169
Recognising Legal Characteristics of the Judgments of the European Court of Justice: Difficult but Not Impossible
Machine learning has improved significantly during the past decades. Computers perform remarkably in formerly difficult tasks. This article reports the preliminary results on the prediction of two characteristics of judgments of the European Court of Justice, which require the knowledge of concepts and doctrines of European Union law and judicial decision-making: The legal importance (doctrinal outcome) and leeway to the national courts and legislators (deference). The analysis relies on 1704 manually labelled judgments and trains a set of classifiers based on word embedding, LSTM, and convolutional neural networks. While all classifiers exceed simple baselines, the overall performance is weak. This suggests first, that the models learn meaningful representations of the judgments. Second, machine learning encounters significant challenges in the legal domain. These arise doe to the small training data, significant class imbalance, and the characteristics of the variables requiring external knowledge. The article also outlines directions for future research.
- 5 December 2022
- Published by IOS Press
- p. 164-169
Open AccessRecognising Legal Characteristics of the Judgments of the European Court of Justice: Difficult but Not Impossible
Machine learning has improved significantly during the past decades. Computers perform remarkably in formerly difficult tasks. This article reports the preliminary results on the prediction of two characteristics of judgments of the European Court of Justice, which require the knowledge of concepts and doctrines of European Union law and judicial decision-making: The legal importance (doctrinal outcome) and leeway to the national courts and legislators (deference). The analysis relies on 1704 manually labelled judgments and trains a set of classifiers based on word embedding, LSTM, and convolutional neural networks. While all classifiers exceed simple baselines, the overall performance is weak. This suggests first, that the models learn meaningful representations of the judgments. Second, machine learning encounters significant challenges in the legal domain. These arise doe to the small training data, significant class imbalance, and the characteristics of the variables requiring external knowledge. The article also outlines directions for future research.
- Open Access
- 1 May 2020
- Published by Cambridge University Press (CUP) in Proceedings of the Design Society: DESIGN Conference
- Vol. 1, 2335-2344
MODULARITY, ROBUSTNESS, AND CHANGE PROPAGATION: A MULTIFACETED RELATION
Increased system robustness is one of the promises of modularity. However, research on the topic has provided conflicting findings. By generating more than 2000 system architectures, this paper shows that the relation between modularity and robustness is multifaceted: Modularity decreases topological robustness, increases robustness to change propagation, and provides economic benefits. Results here confirm the importance of modularity, enable reconciliation of opposing findings from prior research, and guides researchers and practitioners in the selection of appropriate robustness measures.
- 1 May 2020
- Published by Cambridge University Press (CUP) in Proceedings of the Design Society: DESIGN Conference
- Vol. 1, 2335-2344
Open AccessMODULARITY, ROBUSTNESS, AND CHANGE PROPAGATION: A MULTIFACETED RELATION
Increased system robustness is one of the promises of modularity. However, research on the topic has provided conflicting findings. By generating more than 2000 system architectures, this paper shows that the relation between modularity and robustness is multifaceted: Modularity decreases topological robustness, increases robustness to change propagation, and provides economic benefits. Results here confirm the importance of modularity, enable reconciliation of opposing findings from prior research, and guides researchers and practitioners in the selection of appropriate robustness measures.
- 5 September 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Engineering Management
- Vol. 68 (6) , 1867-1879
Process Modularity Over Time: Modeling Process Execution as an Evolving Activity Network
Process modularity describes the extent to which processes can be decomposed into modules to be executed in parallel. So far, research has approached process modularity from a static perspective, not accounting for its temporal evolution. As a result, the understanding of process modularity has been limited to inferences drawn from aggregated analyses that disregard process execution. This article introduces and develops the notion of dynamic process modularity considering the evolving activity network structure as executed by people. Drawing on network science, the article quantifies process modularity over time using archival data from an engineering design process of a biomass power plant. This article shows how studying the temporal evolution of process modularity enables a more complete understanding of activity networks, facilitates the comparison of actual process modularity patterns against formal engineering design stages, and provides data-driven decision-support for process planning and interventions. Finally, managerial recommendations for interface management, resource allocation, and process decomposition are proposed, to help practitioners better to understand and manage dynamic processes.
- 5 September 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Engineering Management
- Vol. 68 (6) , 1867-1879
Process Modularity Over Time: Modeling Process Execution as an Evolving Activity Network
Process modularity describes the extent to which processes can be decomposed into modules to be executed in parallel. So far, research has approached process modularity from a static perspective, not accounting for its temporal evolution. As a result, the understanding of process modularity has been limited to inferences drawn from aggregated analyses that disregard process execution. This article introduces and develops the notion of dynamic process modularity considering the evolving activity network structure as executed by people. Drawing on network science, the article quantifies process modularity over time using archival data from an engineering design process of a biomass power plant. This article shows how studying the temporal evolution of process modularity enables a more complete understanding of activity networks, facilitates the comparison of actual process modularity patterns against formal engineering design stages, and provides data-driven decision-support for process planning and interventions. Finally, managerial recommendations for interface management, resource allocation, and process decomposition are proposed, to help practitioners better to understand and manage dynamic processes.
- Open Access
- 16 October 2018
- Published by Springer Nature in Research in Engineering Design
- Vol. 30 (2) , 251-270
Iterations as the result of social and technical factors: empirical evidence from a large-scale design project
- 16 October 2018
- Published by Springer Nature in Research in Engineering Design
- Vol. 30 (2) , 251-270
Open AccessIterations as the result of social and technical factors: empirical evidence from a large-scale design project