"Multilingual RAG for Corporate Climate Policy Engagement accepted at #EMNLP 2025"

Natural intelligence vs. Artificial intelligence.  Artificial intelligence is a machine language product created by human (natural) intelligence. Thankfully, now a days, intellectual property rights laws (IPR laws) is in practice, what if legislature pass a law on artificial property rights laws (APR laws)?  Let there be a new area of practice. 😊 Renewable Energy vs. Conventional Energy  Save natural resources and make wise use thereof. Natural resources cannot be manufactured. RE is sourced from the Natural resources except wind, solar and hydro.  Air, Water, Noise, Land pollutions is spreading like a cancer, no concrete solution due to dearth of system.   Love the nature, do not spoil it.  Above all is Law of Nature. What you sow, so shall you reap. Fragile eco-system is badly ruined by the human beings resulted in a devastated climate change effect.  Sustainable development is remain in the books only.  Environment and Development is Natural and Neutral permutation-combination, whereas Environment vs. Development is a disaster.  The effect of climate change is horrendous, Country / Nation / City will sink without a prior Notice.  Reduce-Reuse-Recycle is a universal formula to mitigate the pollution effect.

We're also giving a workshop at the Climate Arena conference about 'People power vs AI'. We'll share our experiences with analysing tons of documents: in the Mapping Fossil Ties project we organised 'hackathons' to read through documents we received from universities about their fossil ties with groups of people. We have also used LLMs to help us with our research. Together with moderator Raphaelle Aubert, data journalist at Le Monde, and the workshop participants, we'll discuss pros and cons of both methods and when each is useful in all of our work. Find out more about the programme here: https://lnkd.in/edun9y6D Tickets: https://lnkd.in/ezsjQQZw Arena for Journalism in Europe

🎉 "ReportGRI: Automating GRI Alignment and Report Assessment" has been accepted at the International Conference on Information and Knowledge Management CIKM 2025 in Seoul! The Global Reporting Initiative (#GRI) is the world's most widely used sustainability reporting framework, providing a global common language for organizations to report on their environmental, social, and governance impacts. The Challenge: unstructured corporate reporting makes it hard to: 📊 Compare companies fairly 🔍 Identify reporting gaps ⚖️ Assess disclosure quality #ReportGRI addresses this by automating: ✅ GRI disclosure indexing using advanced NLP ✅ Visual analytics showing topic coverage and gaps ✅ Completeness and materiality assessment powered by LLMs ✅ Cross-company benchmarking capabilities What makes this impactful: • For Sustainability Officers: Ensure GRI compliance before external audits • For Investors: Quickly assess ESG risks and compare portfolio companies • For NGOs & Public: Hold companies accountable with visual evidence 🔓 We've made ReportGRI open-source because we believe sustainability tools should be accessible to all stakeholders working toward a transparent, accountable future. This work represents collaboration between Leuphana University of Lüneburg and University of Zurich's sureal.ai (https://sureal.ai/), bridging the gap between #AI research and real-world #climatepolicy needs. Looking forward to presenting at #CIKM2025 (https://cikm2025.org/) 📄 Paper: https://lnkd.in/eKbqgGrF 🔗 GitHub: https://lnkd.in/ejQH_EGH A big thx goes to: Aida Usmanova, Rana Abdullah, Debayan Banerjee, and Ricardo Usbeck #Sustainability #ESG #GRI #AIforGood #OpenSource #ClimateAction #CorporateTransparency #NLP #CIKM2025 #Research UZH Department of Finance SFI Swiss Finance Institute Swiss Sustainable Finance Building Bridges Florian Esterer Oliver Marchand Dr. Sebastian Gehricke

Anthropic's new article to Context Engineering is a great introduction on how to design better contexts. From the article -- Building with language models is becoming less about finding the right words and phrases for your prompts, and more about answering the broader question of “what configuration of context is most likely to generate our model’s desired behavior?" 🔗 Link in comments.

Super interesting research on hand-curated "Values-Targeted Datasets" can significantly change model behavior that reflects a predetermined set of target values. https://lnkd.in/gHky84YH

This is an excellent and remarkable use case combining the capabilities of deep learning and large language models in oil spill detection and remediation. This system aims to enhance environmental protection efforts and prevent oil spills through informed decision-making and timely intervention. If you are interested in this subject, read further in this article: https://lnkd.in/dHcR7DfM #oilspill #oilspilldetection #oilspillremediation #deeplearning #llms #largelanguagemodels

AGI is an oxymoron. Intelligence is always specific. It abstracts and generalises over contexts, and all abstractions are necessarily biased towards the aims and goals of that particular agency. Intelligence is in fact measured as well as defined by an agency's capacity to adjust to its context(s) in order to achieve that its goals. This insight via the Conference and Labs of the Evaluation Forum (CLEF), an outstanding convocation of people working on a variety of AI-related tasks very related to systems engineering and accountability. I gave a keynote about "coevolving with A(G)I" that included the below "CS 101" slide. The CLEF people were sufficiently actual practitioners to figure out (over the next coffee break) that AI didn't really ever need to know or exhaustively explore everything. Well, it would if it was to be truly omniscient and unbiased, but like the slide says, that's physically impossible. CLEF contributes "to the systematic evaluation of information access systems, primarily through experimentation on shared tasks" with "an independent peer-reviewed conference on a broad range of issues in the fields of multilingual and multimodal information access evaluation, and a set of labs and workshops designed to test different aspects of mono and cross-language Information retrieval systems." I strongly encourage anyone working in data science or language systems to check it out here. Only a small percentage of those participating globally in contests concerning solving the tasks come to the annual meeting, which is always held in Europe. https://lnkd.in/e9PQ673c #AGI #AIEvaluation #systemsAI

Introducing the Next Generation of Conflict Research! BrainBridge, with the support of #KASLebanon, has launched a pioneering initiative that marks a new era in conflict research: the first AI-powered conflict analysis platform in the MENA region. Hate speech online has become one of the most powerful catalysts for violence in the modern era. In #Syria, sectarian rhetoric spread across digital platforms has incited devastating consequences, most recently contributing to the murder of nearly 2,000 Alawites on the Syrian coast in March 2025. The project, “Mapping Syria’s Digital Society: Analysing Discourse on the Alawite Massacres Using AI”, combines political science and advanced machine learning to provide unprecedented insights into how digital discourse fuels real-world conflict. 🚀 Core innovations include: • AI-driven analysis of 100,000+ online reactions • An interactive Knowledge Graph mapping narratives, networks, and rhetoric layers • Public access tools for researchers, journalists, and policymakers worldwide • A comprehensive report and policy briefs offering guidelines for conflict prevention This initiative is a milestone in BrainBridge’s mission to perfect the use of AI and machine learning as tools for world peace. By advancing towards a real-time conflict forecasting platform, BrainBridge is laying the foundations for a future where technology does not merely describe conflicts, but actively helps prevent them. BrainBridge’s vision is bold: to stand at the frontier where AI, political science, and human rights converge, building the frameworks that make global peace not just an aspiration, but an achievable reality. Dr. Talip Al-Khayer #AIForGood #ConflictPrevention #DigitalResearch #AIResearch #Innovation

BrainBridge Launches Groundbreaking AI Project on Conflict Analysis With the support of the Konrad Adenauer Stiftung, I am delighted to announce the launch of a pioneering project that marks a major step forward in applying artificial intelligence to conflict analysis. Developed by BrainBridge, this initiative will harness cutting-edge machine learning techniques to collect and analyse over 100,000 comments and posts on Syria's coastal massacres from social media. The findings will be presented in the form of an interactive knowledge graph, offering the public a dynamic, accessible, and data-driven resource. This freely available platform will serve as a valuable reference for researchers, journalists, policymakers, and anyone seeking to understand the dynamics of conflict and hate speech online. Beyond documenting events, the project represents an essential milestone in pursuing our broader mission: developing AI systems capable of forecasting violence before it occurs, by tracking the evolution of hate speech and its transformation into violent action. The interactive knowledge graph will be released in February next year, with progress updates shared along the way. We warmly welcome questions, collaborations, and especially contributions from those who may hold datasets that could enrich this project. Many thanks to the fantastic technical team who are volunteering their time and effort to make this project a reality Murad Daryousse Dr. Gillian Kant BrainBridge

"Google’s disclosure of the environmental impact of their Gemini models has given rise to a fresh round of debate on the environmental impact of AI and how to measure it." Kasper Groes Albin Ludvigsen digs into the numbers.