Skip to main content

Resultados relacionados

Resultados totales (Incluyendo duplicados): 5
Encontrada(s) 1 página(s)

RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Artículo científico (JournalArticle)

Artificial intelligence reveals unbalanced sustainability domains in funded research

  • Varelas, Panagiotis
  • Larosa, Francesca
  • Contino, Francesco
  • Nerini, Francesco Fuso
  • García-Martínez, Javier
  • Garibo-i-Orts, Óscar
  • Parente, Alessandro
  • Vinuesa, Ricardo
  • Hoyas, Sergio|||0000-0002-8458-7288
  • Conejero, J. Alberto|||0000-0003-3681-7533
[EN] To meet the 2030 Agenda for Sustainable Development, all Sustainable Development Goals (SDGs) must receive adequate and balanced funding. This study applies artificial intelligence to analyze research proposals accepted between 2015 and 2023 in the European Union and the United States, focusing on datasets from the European Research Council and the National Science Foundation, respectively. Despite the growing application of Artificial Intelligence (AI) in various domains, there remains a lack of comprehensive analysis that applies AI to examine funding allocation across SDGs and gender disparities in scientific research. This study addresses this unmet need by using AI to uncover imbalances in funding distribution, offering insights into current funding instruments. We reveal critical coverage disparities across SDGs, with both funding instruments prioritizing SDG 9 (Industry, Innovation, and Infrastructure), highlighting a potential overemphasis on this goal. Additionally, we document pronounced gender imbalances among principal investigators across nearly all SDGs, except for SDG 5 (Gender Equality), in which female researchers are comparatively better represented. Our results indicate an urgent need for more inclusive and balanced approaches to achieve sustainable development, starting with allocation of research funding. By providing a nuanced understanding of funding dynamics and advocating strategic reallocations, this study offers actionable policy design and planning insights to foster a more equitable and comprehensive support system for sustainability-focused research endeavours., PV, AP, and FC acknowledge support from the Energy Transition Fund of Belgium. FL, FFN, and RV acknowledge funding from Digital Futures. FL also acknowledges funding from the European Union's Horizon Europe research and innovation programme under the Marie Sklodowska-Curie grant Agreement No. 101150729. SH is partially funded by project PID2021-128676OB-I00 at Ministerio de Ciencia, Innovation y Universidades/FEDER. JAC was supported by the Generalitat Valenciana, Project PROMETEO CIPROM/2022/21. JGM thanks the funding received from the European Union's Hori-zon 2020 Research and Innovation program under Grant agreement No 872102, the Spanish Ministry of Science, Grant PID2021-1287610B-C21 by MCIN/AEI/10.13039/501100011033 by "ERDF A way of making Europe", the Generalitat Valenciana through CIPROM/2023/2, and the Advanced Materials programme and was supported by MCIN with funding from European Union NextGenerationEU (PRTR-C17.I1) and by Generalitat Valenciana.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Artículo científico (JournalArticle)

Large language models in climate and sustainability policy: limits and opportunities

  • Larosa, F.
  • Garcia-Martinez, J.
  • Fuso-Nerini, F.
  • Vinuesa, R.
  • Hoyas, Sergio|||0000-0002-8458-7288
  • Conejero, J. Alberto|||0000-0003-3681-7533
[EN] Accurate, reliable and updated information support effective decision-making by reducing uncertainty and enabling informed choices. Multiple crises threaten the sustainability of our societies and pose at risk the planetary boundaries, hence requiring usable and operational knowledge. Natural-language processing tools facilitate data collection, extraction and analysis processes. They expand knowledge utilization capabilities by improving access to reliable sources in shorter time. They also identify patterns of similarities and contrasts across diverse contexts. We apply general and domain-specific large language models (LLMs) to two case studies and we document appropriate uses and shortcomings of these tools for two tasks: classification and sentiment analysis of climate and sustainability documents. We study both statistical and prompt-based methods. In the first case study, we use LLMs to assess whether climate pledges trigger cascade effects in other sustainability dimensions. In the second use case, we use LLMs to identify interactions between the sustainable development goals and detects the direction of their links to frame meaningful policy implications. We find that LLMs are successful at processing, classifying and summarizing heterogeneous text-based data helping practitioners and researchers accessing. LLMs detect strong concerns from emerging economies in addressing food security, water security and urban challenges as primary issues. Developed economies, instead, focus their pledges on the energy transition and climate finance. We also detect and document four main limits along the knowledge production chain: interpretability, external validity, replicability and usability. These risks threaten the usability of findings and can lead to failures in the decision-making process. We recommend risk mitigation strategies to improve transparency and literacy on artificial intelligence (AI) methods applied to complex policy problems. Our work presents a critical but empirically grounded application of LLMs to climate and sustainability questions and suggests avenues to further expand controlled and risk-aware AI-powered computational social sciences., FL acknowledges funding from the European Union s Horizon Europe research and innovation programme under the Marie Sklodowska-Curie grant agreement No. 101150729. RV, FFN and FL acknowledge the funding provided by Digital Futures, in their Demonstrator-project program. SH acknowledge the Grant PID2021-128676OB-I00 funded by MCIN/AEI/10.13039/ 501100011033 and by ERDF A Way of Making Europe , by the European Union. JAC acknowledges funding from the program Ayuda a Primeros Proyectos de Investigación (PAID-06-23) from the Vice-rectorate for Research from Universitat Politècnica de València (UPV) and by Generalitat Valenciana, Project PROMETEO CIPROM/2022/21.
RUA. Repositorio Institucional de la Universidad de Alicante
Publication Otros (Other)

Artificial intelligence reveals unbalanced sustainability domains in funded research

  • Varelas, Panagiotis
  • Larosa, Francesca
  • Hoyas, Sergio
  • Conejero, J. Alberto
  • Contino, Francesco
  • Fuso-Nerini, Francesco
  • Garcia-Martinez, Javier
  • Garibo i Orts, Òscar
  • Parente, Alessandro
  • Vinuesa, Ricardo
  • Garcia-Martinez, Javier|||/items/421c2466-c63a-45d7-b799-c59d51faf1bc
To meet the 2030 Agenda for Sustainable Development, all Sustainable Development Goals (SDGs) must receive adequate and balanced funding. This study applies artificial intelligence to analyze research proposals accepted between 2015 and 2023 in the European Union and the United States, focusing on datasets from the European Research Council and the National Science Foundation, respectively. Despite the growing application of Artificial Intelligence (AI) in various domains, there remains a lack of comprehensive analysis that applies AI to examine funding allocation across SDGs and gender disparities in scientific research. This study addresses this unmet need by using AI to uncover imbalances in funding distribution, offering insights into current funding instruments. We reveal critical coverage disparities across SDGs, with both funding instruments prioritizing SDG 9 (Industry, Innovation, and Infrastructure), highlighting a potential overemphasis on this goal. Additionally, we document pronounced gender imbalances among principal investigators across nearly all SDGs, except for SDG 5 (Gender Equality), in which female researchers are comparatively better represented. Our results indicate an urgent need for more inclusive and balanced approaches to achieve sustainable development, starting with allocation of research funding. By providing a nuanced understanding of funding dynamics and advocating strategic reallocations, this study offers actionable policy design and planning insights to foster a more equitable and comprehensive support system for sustainability-focused research endeavours.
RUA. Repositorio Institucional de la Universidad de Alicante
Publication Otros (Other)

Large language models in climate and sustainability policy: limits and opportunities

  • Larosa, Francesca
  • Hoyas, Sergio
  • Conejero, J. Alberto
  • Garcia-Martinez, Javier
  • Fuso-Nerini, Francesco
  • Vinuesa, Ricardo
  • Garcia-Martinez, Javier|||/items/421c2466-c63a-45d7-b799-c59d51faf1bc
Accurate, reliable and updated information support effective decision-making by reducing uncertainty and enabling informed choices. Multiple crises threaten the sustainability of our societies and pose at risk the planetary boundaries, hence requiring usable and operational knowledge. Natural-language processing tools facilitate data collection, extraction and analysis processes. They expand knowledge utilization capabilities by improving access to reliable sources in shorter time. They also identify patterns of similarities and contrasts across diverse contexts. We apply general and domain-specific large language models (LLMs) to two case studies and we document appropriate uses and shortcomings of these tools for two tasks: classification and sentiment analysis of climate and sustainability documents. We study both statistical and prompt-based methods. In the first case study, we use LLMs to assess whether climate pledges trigger cascade effects in other sustainability dimensions. In the second use case, we use LLMs to identify interactions between the sustainable development goals and detects the direction of their links to frame meaningful policy implications. We find that LLMs are successful at processing, classifying and summarizing heterogeneous text-based data helping practitioners and researchers accessing. LLMs detect strong concerns from emerging economies in addressing food security, water security and urban challenges as primary issues. Developed economies, instead, focus their pledges on the energy transition and climate finance. We also detect and document four main limits along the knowledge production chain: interpretability, external validity, replicability and usability. These risks threaten the usability of findings and can lead to failures in the decision-making process. We recommend risk mitigation strategies to improve transparency and literacy on artificial intelligence (AI) methods applied to complex policy problems. Our work presents a critical but empirically grounded application of LLMs to climate and sustainability questions and suggests avenues to further expand controlled and risk-aware AI-powered computational social sciences.
RUA. Repositorio Institucional de la Universidad de Alicante
Publication Otros (Other)

Critical misalignments in climate pledges reveal imbalanced sustainable development pathways

  • Larosa, Francesca
  • Rhomrassi, Lamyae
  • Conejero, J. Alberto
  • Garcia-Martinez, Javier
  • Mallor, Fermín
  • Fuso Nerini, Francesco
  • Vinuesa, Ricardo
  • Garcia-Martinez, Javier|||/items/421c2466-c63a-45d7-b799-c59d51faf1bc
We explore the integration of climate action and Sustainable Development Goals (SDGs) in the first two submissions of nationally determined contributions (NDCs) using an AI-based, human-validated framework. Our goal is to provide ex-ante evidence relevant to assessing policy adequacy. We find disparities in topics of interest with high-income countries emphasizing systemic challenges (health, SDG3) and low-income nations prioritizing the water-energy-food nexus (SDGs 6-7-12) and natural resource management (SDG15). We discuss what these diverging development trajectories imply for the Paris Agreement and the 2030 Agenda for sustainable development in terms of global inequality, sustainable finance flows and multilateral governance.