Pasar al contenido principal

Resultados relacionados

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

DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Artículo científico (JournalArticle)

FoodwebAI: Creating and augmenting food webs with natural language models

  • Noleto-Filho, Eurico M.
  • Keppeler, Friedrich Wolfgang
  • Reis-Filho, José Amorim
  • Giarrizzo, Tommaso
  • Keenlyside, Noel
  • Carvalho, Adriana R.
  • Coll, Marta
  • Steenbeek, Jeroen
  • Angelini, Ronaldo
8 pages, 5 figures, 3 tables, supplementary materials https://doi.org/10.1016/j.softx.2026.102812.-- Data used in this article was collected as part of the federal environmental licensing process associated with the Belo Monte hydroelectric complex (Process # 02001.001848/2006-75).-- Software: https://github.com/Filhomn/FoodWebAI, Food webs illustrate the complex interactions among organisms in ecosystems, crucial for understanding energy and material flows. However, building accurate food webs is challenging due to ecological complexity and data demands. FoodwebAI, a user-friendly web application, addresses this by using AI, specifically OpenAI’s GPT models via the ellmer R package, to enrich species lists with metadata and generate food webs from taxa lists alone. The application offers three functional modules: an AI-powered Food Web Creation module that generates trophic structure from a species list alone; an Ecopath module that allows users to visualize and augment food webs exported from Ecopath with Ecosim, including AI-generated icons for nodes; and a Manual module for exploratory food web construction. While still experimental, FoodwebAI demonstrates how AI can support ecological research by improving data accessibility and understanding of complex networks. The app is available at https://foodwebai.shinyapps.io/foodweb under a General Public License (GPLv3), JARF, TG and FWK acknowledge financial support from FADESP. TG is CNPq productivity fellowship (Grants 308528/2022–0). MC and JS acknowledge financial support from the Spanish project ProOceans (PID2020-118097RB-I00, Ministerio de Ciencia e Innovación, Proyectos de I + D + I), Oceanic (PID2024-162762OB-I00 financiado por MCIN/AEI/10.13039/501100011033/FEDER, UE) and MOIRAI (European Union, Grant Agreement No. 101180994). MC acknowledges institutional support of the ’Severo Ochoa Center of Excellence‘ accreditation (CEX2019–000928-S). FWK acknowledges CNPq for financial support through a Bolsa Conhecimento Brasil (BCB) fellowship (grant no. 314782/2025-7). This study received support from Norte Energia S.A. under the ANEEL Development and Innovation Program (PD-07427-0325/2025). EMNF acknowledges funding from the European Union’s Horizon Europe research and innovation programme under the Marie Skłodowska-Curie grant agreement No. 101153695, Peer reviewed