Archivio istituzionale della ricerca - Alma Mater Studiorum Università di Bologna
Archivio istituzionale della ricerca - Alma Mater Studiorum Università di BolognaNot a member yet
336958 research outputs found
Sort by
Counting rather than weighing: metrological analysis and machine learning reveal the monetary potential of pre-contact Ecuadorian axe-monies
No full textThis article investigates the potential monetary function of axe-monies from pre-contact Ecuador (500–1532 CE), a widely diffused and morphologically consistent copper-alloy artifact. Drawing on a dataset of 3,588 specimens, we employ a multidisciplinary approach combining metrological analysis, computer vision, and machine learning techniques to evaluate the presence of weight-based or dimensional standardization and morphological regularities. Our findings challenge the hypothesis that these objects were regulated by weight, as no metrological clusters emerge from the data. Instead, we identify two distinct dimensional classes and a high degree of typological consistency, suggesting intentional standardization based on form rather than mass, with triangular axes being the smallest and lightest, and those with a broad cutting edge and quadrangular heel the largest and heaviest.These results support the interpretation of Ecuadorian axe-monies as fiduciary currency, counted rather than weighed, and contribute to broader discussions on the emergence of money, measurement systems, and economic behaviour in pre-modern societies
INTRODUCTION. Policy Advisory Systems: Research Agendas and Comparative Approaches
No full textMany challenges are associated with developing, delivering, and using policy advice. This is equally true for researching it. These challenges have intensified as modern governance and has become more complex and as new advisory supplies and practices join an already dizzying array of considerations. Recognizing this, policy scholars have turned to studying policy advice in more systematic terms. That is, as policy advisory systems (PAS), or the assemblage of formal and informal advisory units and practices, inside and outside of government, that exist at a given time and with which governments and other actors engage for policymaking purposes (Craft and Halligan 2020). This has facilitated moving from a traditional focus on discreet sets of advisers, especially civil services, to a more synergistic focus that recognizes the pluralism and interactive character of advisory activit
Benchmarking in Neuro-Symbolic AI
No full textNeural-symbolic (NeSy) AI has gained a lot of popularity by enhancing learning models with explicit reasoning capabilities. Both new systems and new benchmarks are constantly introduced and used to evaluate learning and reasoning skills. The large variety of systems and benchmarks, however, makes it difficult to establish a fair comparison among the various frameworks, let alone a unifying set of benchmarking criteria. This paper analyzes the state-of-the-art in benchmarking NeSy systems, studies its limitations, and proposes ways to overcome them. We categorize popular neural-symbolic frameworks into three groups: model-theoretic, proof-theoretic fuzzy, and proof-theoretic probabilistic systems. We show how these three categories have distinct strengths and weaknesses, and how this is reflected in the type of tasks and benchmarks to which they are applied
LIP-CAR: A Learned Inverse Problem Approach for Medical Imaging with Contrast Agent Reduction
No full textCirculating Populist Sentiments in 21st Century Film and TV Fiction in Italy
No full textThe aim of this project is to study the role played by fictional audiovisual media in cultures of populism, in 21st-century Italy. In past decades, liberal democracies have been threatened by the rise of populist movements, hegemonizing complex social and economic crises through a pattern of recurring, simplified narratives. Key features of this “populist imagination” are: the rhetorical construction and idealization of national “people”; its habitual relationship to a charismatic leader; a social binary between that “people” and a corrupt elite or external enemy; and contempt for traditional political processes. The role of the media in the affirmation of populism has been widely acknowledged: political narratives and audiovisual fiction spread through an interrupted flow in the mediascape. Though much research has been dedicated to the role of the information media, little work has illustrated the extent to which fictional narratives are a crucial actant in the populist mediascape
Legal Lay Summarization: Exploring Methods and Data Generation with Large Language Models
Full text linkThis paper explores advancements in Natural Language Processing (NLP) for legal lay summarization by systematically analyzing existing methodologies, datasets, and research findings. We review current literature, highlighting key challenges such as data scarcity and the complexity of legal language. A primary contribution of this study is the development of LegalEase, a specialized dataset designed to improve model training for summarizing legal documents in layman’s terms. Our findings demonstrate that subdomain-specific datasets within the legal domain outperform general legal datasets in enhancing NLP model performance for generating accurate and comprehensible legal summaries. The insights and methodologies presented provide a foundation for future research in legal lay summarization
Approaching the AI Act... with AI: LLMs and knowledge graphs to extract and analyse obligations
No full textThe EU Artificial Intelligence Act (AIA) exemplifies the growing complexity of digital regulation in the domain of computer technologies. Characterised by abstract terminology, multi-layered provisions, and intersecting regulatory requirements, the AIA poses significant challenges for the identification and interpretation of legal obligations, making compliance a demanding and potentially error-prone endeavour for legal professionals and organisations alike. Recent advances in Artificial Intelligence (AI), particularly in the fields of Natural Language Processing (NLP) and Large Language Models (LLMs), offer promising support for addressing these challenges. By automating the extraction and structuring of legal rules, AI-based tools have the potential to assist regulatory compliance activities and provide more systematic insights into complex legislative frameworks. This paper presents an experiment combining NLP techniques and LLMs to automate the extraction and structuring of legal obligations from the AIA. The approach is based on a modular workflow comprising four main stages: identification of obligations, filtering of deontic statements, analysis of deontic content, and the construction of searchable knowledge graphs. The experiment employed the LLaMA 3.3 70B model, supported by more traditional NLP tools. Five experts (4 Ph.D. students and 1 post-doc in legal informatics and philosophy) evaluated the system’s performance on a subset of cases. The results indicate a precision of 93% in the obligation filtering phase and over 99% accuracy in classifying obligation types, addressees, and predicates. A quantitative analysis of the extracted and analysed obligations revealed a predominance of prescriptive obligations (603 out of 729 total), among which 136 are imposed on the European Commission, while 88 consist of informative duties. The results are in line with current discussions around the AI Act regulatory approach. These findings underscore the potential of LLM-based tools to enhance regulatory compliance and analysis. Future research will focus on extending the system to additional EU regulations and integrating formal ontologies to enable more advanced representations of legal obligations
Preface of "Methods and Applications for Modeling and Simulation of Complex Systems"
No full textThis book constitutes the refereed proceedings of the 24th Asia Simulation Conference on Methods and Applications for Modeling and Simulation of Complex Systems, AsiaSim 2025, held in Singapore, during November 17-19, 2025. The 19 full papers and 8 short papers presented here, were carefully reviewed and selected from 54 submissions. They are grouped into the following topics: Agent Based Simulation, Simulation Methods and Tools, Visualization, Modeling Methodology, and Simulation Applications in Science and Engineering
A transformer-based approach for source code classification for heterogeneous device mapping
Full text linkThe optimization of code allocation for heterogeneous architectures, such as Central Processing Units (CPUs) and Graphics Processing Units (GPUs), remains challenging due to the limitations of traditional compiler heuristics and existing machine learning approaches. This paper presents a systematic evaluation of Large Language Models (LLMs) for classifying source code execution targets in heterogeneous device mapping. We fine-tune and compare six models: Distilled Bidirectional Encoder Representations from Transformers (DistilBERT), Code Bidirectional Encoder Representations from Transformers (CodeBERT), Code Bidirectional Encoder Representations from Transformers with RoBERTa (Robustly Optimized BERT Pretraining Approach) architecture (CodeBERTa), CodeT5, jTrans, and Deep Learning Low Level Virtual Machine (DeepLLVM), trained on Open Computing Language (OpenCL) kernels. Results show that general-purpose LLMs achieve up to 92.8% accuracy, matching or surpassing code-specific models, and outperform the previous state of the art (DeepLLVM) by up to 5%. Our findings indicate that LLMs pre-trained on general text are not necessarily inferior to code-specialized models, with tokenizer design and pre-training objectives impacting performance more than domain specialization. These results demonstrate the effectiveness of Transformer-based LLMs as a state-of-the-art approach for source code classification in heterogeneous computing contexts
Nanofiltration efficiency in the purification of lactose from ultrafiltered acid whey
No full textThe recovery of lactose from acid whey has not yet been implemented at an industrial scale, primarily due to high concentration of lactic acid and minerals. This study demonstrates the feasibility of purifying and concentrating lactose derived from the permeate of ultrafiltered acid whey. A simple batch nanofiltration process was developed to achieve simultaneous deacidification and demineralization by integrating a concentration step with a subsequent constant-volume diafiltration step. The process enables efficient lactose recovery and provides an aqueous by-product containing primarily lactic acid and monovalent salts. Experiments were conducted with real solutions consisting mainly of 30 g/L lactose, 4.3 g/L lactic acid and 2.5 g/L electrolytes. Commercial spiral wound polyamide membranes were tested at pH 4.0 and 50 °C, and membrane performance was assessed based on rejection measures and a preliminary quantification of fouling. The diafiltration step was essential to ensure a final product with a lactose concentration close to 100 g/L and a lactic acid to lactose ratio not exceeding 0.030 g/g. Process efficiency was determined by the optimal combination of the transmembrane fluxes imposed in both steps: high lactose purity (95%), yield (99.6%) and demineralization efficiency (86%) were obtained when operating at fluxes at 30 L/(m2h) with a minimum pumping requirement of 3.1 kWh/m3 of feed. However, while increasing the transmembrane flux improved performance and energy efficiency, a significant increase in water demand was observed in the diafiltration step. Nevertheless, water recovery via an additional reverse osmosis step accounts for no more than 20 % of the total energy requirement