Thailand

This paper introduces and evaluates ChatNetZero, a large-language model (LLM) chatbot developed through Retrieval-Augmented Generation (RAG), which uses generative AI to produce answers grounded in verified, climate-domain specific information. We describe ChatNetZero&#39;s design, particularly the innovation of anti-hallucination and reference modules designed to enhance the accuracy and credibility of generated responses. To evaluate ChatNetZero&#39;s performance against other LLMs, including GPT-4, Gemini, Coral, and ChatClimate, we conduct two types of validation: comparing LLMs&#39; generated responses to original source documents to verify their factual accuracy, and employing an expert survey to evaluate the overall quality, accuracy and relevance of each response. We find that while ChatNetZero responses show higher factual accuracy when compared to original source data, experts surveyed prefer lengthier responses that provide more context. Our results highlight the importance of prioritizing information presentation in the design of domain-specific LLMs to ensure that scientific information is effectively communicated, especially as even expert audiences find it challenging to assess the credibility of AI-generated content.

ACL 2024

Evaluating ChatNetZero, an LLM-Chatbot to Demystify Climate Pledges

net zero pledges

retrieval-augmented generation

large language models

natural language processing

climate change

### Welcome to The 63rd Annual Meeting of the Association for Computational Linguistics (ACL 2025)

Message from the General Chair: 
*It is my great pleasure and honor to welcome you to the 63rd Annual Meeting of the Association for Computational Linguistics (ACL 2025), held in beautiful Vienna, Austria, from July 27 to August 1, 2025. ACL2025continues our field’s tradition of excellence in scholarship, innovation, and inclusivity, and I am deeply grateful to the many volunteers who have worked tirelessly to bring this event to life.* 
[Read more](https://drive.google.com/file/d/1GI_hvOpjswAuYdUTromfeDiPpCcqidwg/view?usp=sharing)

To access this event page, you need to log in with the **email address you registered with**. Access credentials will be sent to your email from Underline - subject line "Welcome to ACL 2025". Please be sure to check your spam email folder if you do not see an email confirmation right away.

Please log in

To access this event page, you need to log in with the **email address you registered with**. 

ACL 2025

Welcome to The 63rd Annual Meeting of the Association for Computational Linguistics (ACL 2025)

This paper introduces and evaluates ChatNetZero, a large-language model (LLM) chatbot developed through Retrieval-Augmented Generation (RAG), which uses generative AI to produce answers grounded in verified, climate-domain specific information. We describe ChatNetZero's design, particularly the innovation of anti-hallucination and reference modules designed to enhance the accuracy and credibility of generated responses. To evaluate ChatNetZero's performance against other LLMs, including GPT-4, Gemini, Coral, and ChatClimate, we conduct two types of validation: comparing LLMs' generated responses to original source documents to verify their factual accuracy, and employing an expert survey to evaluate the overall quality, accuracy and relevance of each response. We find that while ChatNetZero responses show higher factual accuracy when compared to original source data, experts surveyed prefer lengthier responses that provide more context. Our results highlight the importance of prioritizing information presentation in the design of domain-specific LLMs to ensure that scientific information is effectively communicated, especially as even expert audiences find it challenging to assess the credibility of AI-generated content.

workshop paper

### Welcome!
The 62nd Annual Meeting of the Association for Computational Linguistics (ACL 2024) will take place in Bangkok, Thailand from August 11th to 16th, 2024. Our Virtual Poster Sessions will take place online Thursday, August 22, 2024.

You are required to register for this event. **Please register [here](https://2024.aclweb.org/registration). **

If you have already registered, please check your inbox for an email from Underline granting you access to ACL 2024 content.

Please register!

The 62nd Annual Meeting of the Association for Computational Linguistics (ACL 2024) will take place in Bangkok, Thailand from August 11th to 16th, 2024. More information will be announced soon.

To better understand how extreme climate events impact society, we need to increase the availability of accurate and comprehensive information about these impacts. We propose a method for building large-scale databases of climate extreme impacts from online textual sources, using LLMs for information extraction in combination with more traditional NLP techniques to improve accuracy and consistency. We evaluate the method against a small benchmark database created by human experts and find that extraction accuracy varies for different types of information. We compare three different LLMs and find that, while the commercial GPT-4 model gives the best performance overall, the open-source models Mistral and Mixtral are competitive for some types of information.

Using LLMs to Build a Database of Climate Extreme Impacts

Climate communication is often seen by the NLP community as an opportunity for machine translation, applied to ever smaller languages. However, over 90% the world's linguistic diversity comes from languages with `primary orality' and mostly spoken in non-Western oral societies. A case in point is the Aboriginal communities of Northern Australia, where we have been conducting workshops on climate communication, revealing shortcomings in existing communication practices along with new opportunities for improving intercultural communication. We present a case study of climate communication in an oral society, including the voices of many local people, and draw several lessons for the research program of NLP in the climate space.

Envisioning NLP for intercultural climate communication

Across countries, a noteworthy paradigm shift towards a more sustainable and environmentally responsible economy is underway. However, this positive transition is accompanied by an upsurge in greenwashing, where companies make exaggerated claims about their environmental commitments. To address this challenge and protect consumers, initiatives have emerged to substantiate green claims. With the proliferation of environmental and scientific assertions, a critical need arises for automated methods to detect and validate these claims at scale. In this paper, we introduce EnClaim, a transformer network architecture augmented with stylistic features for automatically detecting claims from open web documents or social media posts. The proposed model considers various linguistic stylistic features in conjunction with language models to predict whether a given statement constitutes a claim. We have rigorously evaluated the model using multiple open datasets. Our initial findings indicate that incorporating stylistic vectors alongside the BERT-based language model enhances the overall effectiveness of environmental claim detection.

EnClaim: A Style Augmented Transformer Architecture for Environmental Claim Detection

Although food consumption represents a sub- stantial global source of greenhouse gas emis- sions, assessing the environmental impact of off-the-shelf products remains challenging. Currently, this information is often unavailable, hindering informed consumer decisions when grocery shopping. The present work introduces a new set of models called LEAF, which stands for Linguistic Environmental Analysis of Food Products. LEAF models predict the life-cycle environmental impact of food products based on their name. It is shown that LEAF models can accurately predict the environmental im- pact based on just the product name in a multi- lingual setting, greatly outperforming zero-shot classification methods. Models of varying sizes and capabilities are released, along with the code and dataset to fully reproduce the study.

LEAF: Predicting the Environmental Impact of Food Products based on their Name

In this study, we explore the use of Large Language Models (LLMs) such as GPT-4 to extract and analyze the latent narrative messaging in climate change-related news articles from North American and Chinese media. By defining "narrative messaging" as the intrinsic moral or lesson of a story, we apply our model to a dataset of approximately 15,000 news articles in English and Mandarin, categorized by climate-related topics and ideological groupings. Our findings reveal distinct differences in the narrative values emphasized by different cultural and ideological contexts, with North American sources often focusing on individualistic and crisis-driven themes, while Chinese sources emphasize developmental and cooperative narratives. This work demonstrates the potential of LLMs in understanding and influencing climate communication, offering new insights into the collective belief systems that shape public discourse on climate change across different cultures.

Large Scale Narrative Messaging around Climate Change: A Cross-Cultural Comparison

Following the introduction of the European Sustainability Reporting Standard (ESRS), companies will have to adapt to a new policy and provide mandatory sustainability reports. However, implementing such reports entails a challenge, such as the comprehension of a large number of textual information from various sources. This task can be accelerated by employing Large Language Models (LLMs) and ontologies to effectively model the domain knowledge. 

In this study, we extended an existing ontology to model ESRS Topical Standard for disclosure. The developed ontology would enable automated reasoning over the data and assist in constructing Knowledge Graphs (KGs). Moreover, the proposed ontology extension would also help to identify gaps in companies' sustainability reports with regard to the ESRS requirements. Additionally, we extracted knowledge from corporate sustainability reports via LLMs guided with a proposed ontology and developed their KG representation.

Structuring Sustainability Reports for Environmental Standards with LLMs guided by Ontology

Misinformation regarding climate change is a key roadblock in addressing one of the most serious threats to humanity. This paper investigates factual accuracy in large language models (LLMs) regarding climate information. Using true/false labeled Q&A data for fine-tuning and evaluating LLMs on climate-related claims, we compare open-source models, assessing their ability to generate truthful responses to climate change questions. We investigate the detectability of models intentionally poisoned with false climate information, finding that such poisoning may not affect the accuracy of a model's responses in other domains. Furthermore, we compare the effectiveness of unlearning algorithms, fine-tuning, and Retrieval-Augmented Generation (RAG) for factually grounding LLMs on climate change topics. Our evaluation reveals that unlearning algorithms can be effective for nuanced conceptual claims, despite previous findings suggesting their inefficacy in privacy contexts. These insights aim to guide the development of more factually reliable LLMs and highlight the need for additional work to secure LLMs against misinformation attacks.

Unlearning Climate Misinformation in Large Language Models

Aligning unstructured climate policy documents according to a particular classification taxonomy with little to no labeled examples is challenging and requires manual effort of climate policy researchers. In this work we examine whether large language models (LLMs) can act as an effective substitute or assist in the annotation process. Utilizing a large set of text spans from Paris Agreement Nationally Determined Contributions (NDCs) linked to United Nations Sustainable Development Goals (SDGs) and targets contained in the Climate Watch dataset from the World Resources Institute in combination with our own annotated data, we validate our approaches and establish a benchmark for model performance evaluation on this task. With our evaluation benchmarking we quantify the effectiveness of using zero-shot or few-shot prompted LLMs to align these documents.

Aligning Unstructured Paris Agreement Climate Plans with Sustainable Development Goals

Climate change poses an urgent global problem that requires efficient data analysis mechanisms to provide insights into climate-related discussions on social media platforms. This paper presents a framework aimed at understanding social media users' perceptions of various climate change topics and uncovering the insights behind these perceptions. Our framework employs large language model to develop a taxonomy of factual claims related to climate change and build a classification model that detects the truthfulness stance of tweets toward the factual claims. The findings reveal two key conclusions: (1) The public tends to believe the claims are true, regardless of the actual claim veracity; (2) The public shows a lack of discernment between facts and misinformation across different topics, particularly in areas related to politics, economy, and environment.

Granular Analysis of Social Media Users’ Truthfulness Stances Toward Climate Change Factual Claims

In the natural sciences, a common form of scholarly document is a physical sample record, which provides categorical and textual metadata for specimens collected and analyzed for scientific research. Physical sample archives like museums and repositories publish these records in data repositories to support reproducible science and enable the discovery of physical samples. However, the success of resource discovery in such interfaces depends on the completeness of the sample records. We investigate approaches for automatically completing the scientific metadata fields of sample records. We apply large language models in zero and few-shot settings and incorporate the hierarchical structure of the taxonomy. We show that a combination of record summarization, bottom-up taxonomy traversal, and few-shot prompting yield F1 as high as 0.928 on metadata completion in the Earth science domain.

Downloads

Next from ACL 2024

Using LLMs to Build a Database of Climate Extreme Impacts

Similar lecture

Question answering in Climate Adaptation: Model Development and Evaluation with Expert Feedback

.css-70qvj9{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;}Downloads

Next from ACL 2024

Using LLMs to Build a Database of Climate Extreme Impacts

Similar lecture

Question answering in Climate Adaptation: Model Development and Evaluation with Expert Feedback

Downloads