The new ESG-RC model, to be released in October 2023, is built upon our TRAIN Platform and delivers Root Cause Analysis to empower investors and enterprises to navigate the complexities of ESG, championing sustainable business and responsible investment.

PRINCETON, N.J., Sept. 19, 2023 /PRNewswire/ — With over 1,800,000 downloads, Parabole.ai’s ESG-BERT model, was one of, if not the earliest ESG-trained language models to be released to the general public. Parabole.ai extends its commitment to open source and pioneering advancements in Causal Intelligence and ESG-focused AI. Enriched through a comprehensive dataset including sustainability reports and pivotal ESG materials, ESG-BERT is trained to discover intricate ESG contexts with industry leading precision. In a benchmark study of ESG-BERT, the standard BERT model scored a F1 rating of 0.79, while the ESG-BERT model delivered a score of 0.90.

Building on this legacy, the forthcoming ESG-RC model promises enhanced capabilities, integrating powerful Root Cause Analysis. This unique attribute is tailored to guide users in understanding the core dynamics behind ESG risks and opportunities. The innovation within ESG-RC extends the power and scope of the original model, spotlighting critical supply chain concerns, including new insights into Modern Day Slavery and Global Food Supply Chain practices and violations. While ESG-RC will be released in October, ESG-BERT will continue to be available here: ESG-BERT Bitbucket.

Enhanced Features of ESG-RC:

• Root Cause Analysis: discovers the cause and effect of core ESG risks and opportunities. Discovers the “why?” behind the statements, claims and outcomes regarding sustainability reporting.

• Text Classification: Seamless sorting of text into ESG components, aiding in risk and opportunity identification.
• Contextual Analysis: Comprehensive exploration of ESG narratives, highlighting areas requiring detailed scrutiny.
• Entity Extraction: Identification of essential entities, such as corporate names or ESG metrics, fostering knowledge graphs and longitudinal performance tracking.
• Risk Analysis: Advanced tools for a thorough ESG risk evaluation, showcasing latent opportunities.
• Reporting: Transparent, detailed report generation, presenting ESG metrics to stakeholders with clarity.

Anthony Sarkis, Parabole.ai’s Head of Strategy, reflected on Parabole’s history in ESG AI development: “The TRAIN Platform provides our enterprise clients with a toolkit to craft knowledge-enriched causal models, as well as a powerful library of frameworks to solve novel and complex problems across manufacturing with supply chain sustainability a key focus. In 2020, we were pioneers in the Generative AI field with the launch of our open-source ESG-BERT model. Its broad acceptance attests to our team’s proficiency in developing language models that seamlessly complement our Causal Platform. The upcoming ESG-RC model, set to define root cause analysis in ESG, will be open-source—supporting our dedication to reinforcing global sustainability efforts for all.”

Our TRAIN platform can create custom frameworks that extend our growing library of Causal Frameworks, tailored for manufacturing enterprises, government agencies, and standards organizations. These solutions are designed to amplify insights and decision-making capabilities around all supply chain and sustainability vectors. Please contact Jonathan Doan, Vice President at jd@parabole.ai to learn how TRAIN can empower your enterprise.

About Parabole.ai

At Parabole.AI, we are committed to advancing the positive impact of artificial intelligence and knowledge-driven decision-making. Our mission is centered on the strategic fusion of knowledge and data—an approach that enables enterprises to navigate the intricate and ever-evolving landscape of today’s global challenges.

View original content to download multimedia:https://www.prnewswire.com/news-releases/paraboleai-announces-the-release-of-esg-rc-an-open-source-follow-up-to-the-popular-esg-bert-advanced-esg-model-301931734.html

SOURCE Parabole.ai