Dalla cattura del carbonio ai big data: nuove pubblicazioni evidenziano la ricerca della comunità del Dipartimento di Eccellenza L EMbeDS della Scuola Sant Anna

From carbon capture to big data: new publications highlight research by the L EMbeDS, one of the Departmens of Excellence of the Sant Anna School g.cei Tue, 12/10/2024 - 16:22 The research conducted by the L’EMbeDS, one of the Department of Excellence of the Sant Anna School, community continues to grow and receive recognition through publications in high-profile international journals and conference proceedings. Among the latest publications are studies on diverse and innovative topics, ranging from the impacts of different carbon removal technologies, to the use of big data in small and medium enterprises, to methodological developments concerning network and functional data analysis.In the article, The public use of early-stage scientific advances in carbon dioxide removal: a science-technology-policy-media perspective, recently appeared in the prestigious journal Environmental Research Letters, Giorgio Tripodi (a prior L’EMbeDS postdoc, now at Northwestern University in Chicago), Fabrizio Lillo (Scuola Normale Superiore and University of Bologna) and Roberto Mavilia (Bocconi University), together with Andrea Mina, Francesca Chiaromonte e Francesco Lamperti (three Sant’Anna faculty and members of L’EMbeDS) investigated and compared different Carbon Dioxide Removal solutions (CDRs). CDRs are essential to meet Paris Agreement objectives and curb climate change, yet their maturity and current ability to operate at scale are highly debated. The rapid development, deployment, and diffusion of these techniques will likely require the coordination of science, technology, policy, and societal support. The article adopts a bibliometric approach to quantify the public use of early-stage research in CDR. Using statistical techniques, the authors evaluate the impacts of advances in eight different CDRs on (i) further production of scientific knowledge, (ii) technological innovation, and (iii) policy and media discussion. The main result is that advances in CDR are of significant social value and generate positive, yet heterogeneous spillovers within science and from science to technology, policy, and media. In particular, advances in Direct Air Capture spur further research and tend to result in patentable technologies, while Blue Carbon and Bio-energy with Carbon Capture and Storage appear to gain relative momentum in the policy and public debate. Moreover, scientific production and collaborations tend to cluster geographically by type of CDR, potentially affecting long-term carbon removal strategies. Overall, findings suggest the existence of coordination gaps between science, technology, policy, and public support. Importantly, the study was also instrumental to research conducted as part of Francesco Lamperti’s ERC award FIND.In the article What “V” of the big data influence SMEs’ open innovation breadth and depth? An empirical analysis, recently appeared in the high impact R D Management journal, Giulio Ferrigno, Saverio Barabuffi, Enrico Marcazzan e Andrea Piccaluga (postdocs and faculty at Sant’Anna, and members of L’EMbeDS), tackle the topic of Open Innovation (OI), a paradigm that has garnered attention in recent years. Specifically, the study explores how three “V”s characterizing the use of Big Data -- namely, Volume, Velocity and Variety -- affect the OI research of small and medium enterprises (SMEs). Issues related to the use of Big Data have been often examined in the context of high-tech firms, but their effects on SMEs and their OI research strategies have not been extensively studied. The authors address this gap with a quantitative analysis on a sample of 123 Italian SMEs. The findings suggest that, in Italian SMEs, Big Data significantly affect OI breadth, leading to increased external collaborations, but not OI depth. Moreover, the impact varies among the different “V”s of Big Data, with some characteristics having a more pronounced effect on SMEs’ OI strategies. Drawing on these insights, the article sheds light on the interplay between Big Data, with its characteristics, and SMEs’ OI research, and proposes avenues for further research and practice. The article Efficient Network Embedding by Approximate Equitable Partitions, co-authored by Giuseppe Squillace and Mirco Tribastone (both at IMT Lucca), Max Tschaikowski (Aalborg University, Denmark) and Andrea Vandin (a faculty at Sant’Anna and a member of L’EMbeDS), exemplifies collaborations between Tuscan research institution and international partners. It recently appeared in the proceedings of the prestigious IEEE International Conference on Data Mining (ICDM), a major venue in computer science -- considered the “world’s premier research conference in Data Mining”. In many domains of science and engineering, complex information is represented through networks -- where nodes are the studied entities and edges are relations among them. This structured information allows one, for instance, to identify the ‘most important’ nodes, e.g., individuals that dominate communication in a social network. The article proposes an innovative technique to identify similar nodes in a network, enabling novel analyses, and, most importantly, allowing researchers to reduce large networks into smaller ones while preserving fundamental network properties. This can be used to tame the complexity stemming from ever growing amount of data. The authors compare their method against state-of-the-art techniques on benchmark networks, reporting superior precision and scalability for several analysis tasks of interest. Notably, the analysis speedup can be as much as 3 orders of magnitude (1000 times faster); this allows one to handle very large networks, which could not be efficiently analyzed using most of existing techniques. The article FAStEN: An Efficient Adaptive Method for Feature Selection and Estimation in High-Dimensional Functional Regressions, co-authored by Tobia Boschi (IBM Research Dublin), Lorenzo Testa (a prior Sant’Anna allievo, now pursuing a PhD at Carnegie Mellon University in Pittsburgh), Francesca Chiaromonte (a faculty at Sant’Anna and Penn State University and the scientific coordinator of L’EMbeDS) and Matthew Reimherr (Amazon and Penn State), provides another example of collaborations between Tuscan research institution and international partners. It recently appeared in the Journal of Computational and Graphical Statistics – one of the most recognized venues for research in computational statistics. FAStEN is a new method for efficiently selecting and estimating important features in high-dimensional regression models, where both predictors and responses are represented as functions, i.e. curves or surfaces on continuous domains. FAStEN is designed to handle complex datasets, such as brain fMRI data, where it is essential to identify the most relevant variables from among thousands of potential functional predictors. Combining functional principal components with an advanced optimization algorithm, FAStEN significantly reduces computation time without sacrificing accuracy. The article describes an application to brain fMRI seeking potential associations between brain activity and heart rate variability, FAStEN identified regions involved in heart rate control and emotional anticipation. This new tool holds promise for the analysis of functional data that are increasingly ubiquitous in a variety of fields -- from neurosciences, to epidemiology and public health, to economics and finance. The main image is an AI-generated image created by Microsoft Copilot Inserisci negli avvisi in Homepage Spento Data pubblicazione Tue, 12/10/2024 - 16:02 Soggetto Embeds Titolo card From carbon capture to big data: new publications highlight research by the L EMbeDS, one of the Departmens of Excellence of the Sant Anna School Sommario Recent studies explore innovative topics and receive recognition through publications in high-profile international journals and conference proceedings

Read on the original site