论文检索

Social Aspects · Fairness, Equity, Justice and Safety

Shicheng Xu, Liang Pang, Yunchang Zhu, Huawei Shen, Xueqi Cheng

Vision-language alignment in Large Vision-Language Models (LVLMs) successfully enables LLMs to understand visual input. However, we find that existing vision-language alignment methods fail to transfer the existing safety mechanism for text in LLMs to vision, which leads to vulnerabilities in toxic image. To explore the cause of this problem, we give the insightful explanation of where and how the safety mechanism of LVLMs operates and conduct comparative analysis between text and vision. We find that the hidden states at the specific transformer layers play a crucial role in the successful activation of safety mechanism, while the vision-language alignment at hidden states level in current methods is insufficient. This results in a semantic shift for input images compared to text in hidden states, therefore misleads the safety mechanism. To address this, we propose a novel Text-Guided vision-language Alignment method (TGA) for LVLMs. TGA retrieves the texts related to input vision and uses them to guide the projection of vision into the hidden states space in LLMs. Experiments show that \textbf{TGA} not only successfully transfers the safety mechanism for text in basic LLMs to vision in vision-language alignment for LVLMs without any safety fine-tuning on the visual modality but also maintains the general performance on various vision tasks (Safe and Good). Code is in supplemental material and will be released on GitHub after acceptance.

Miscellaneous Aspects of Machine Learning · Causality

Anpeng Wu, Haiyi Qiu, Zhengming Chen, Zijian Li, Ruoxuan Xiong, Fei Wu, Kun Zhang

Networked interference, also known as the peer effect in social science and spillover effect in economics, has drawn increasing interest across various domains. This phenomenon arises when a unit’s treatment and outcome are influenced by the actions of its peers, posing significant challenges to causal inference, particularly in treatment assignment and effect estimation in real applications, due to the violation of the SUTVA assumption. While extensive graph models have been developed to identify treatment effects, these models often rely on structural assumptions about networked interference, assuming it to be identical to the social network, which can lead to misspecification issues in real applications. To address these challenges, we propose an Interference-Agnostic Causal Graph Transformer (CauGramer), which aggregates peers information via $L$-order Graph Transformer and employs cross-attention to infer aggregation function for learning interference representations. By integrating confounder balancing and minimax moment constraints, CauGramer fully incorporates peer information, enabling robust treatment effect estimation. Extensive experiments on two widely-used benchmarks demonstrate the effectiveness and superiority of CauGramer. The code is available at https://github.com/anpwu/CauGramer.

Social Aspects · Accountability, Transparency and Interpretability

Yazheng Liu, Sihong Xie

Graphs are ubiquitous in social networks, chemical molecules, and financial data, where Graph Neural Networks (GNNs) achieve superior predictive accuracy. Graphs can beevolving, while understanding how GNN predictions respond to the evolution provides significant insight and trust. We explore the problem of explaining evolving GNN predictions due to continuously changing edge weights.We introduce a layer edge-based explanation to balanceexplanation fidelity and interpretability.We propose a novel framework to address the challenges of axiomatic attribution and the entanglement of multiple computational graph paths due to continuous change of edge weights. We first design an axiomatic attribution of the evolution of the model prediction to message flows, then develop Shapley value to fairly map message flow contributions to layer edges.We formulate a novel optimization problem to find the critical layer edges based on KL-divergence minimization. Extensive experiments on eight datasets for node classification, link prediction, and graph classification tasks with evolving graphs demonstrate the better fidelity and interpretability of the proposed method over the baseline methods. The code is available at https://github.com/yazhengliu/Axiomatic-Layer-Edges/tree/main.

Social Aspects · Fairness, Equity, Justice and Safety

Joonsung Jeon, Woo Jae Kim, Suhyeon Ha, Sooel Son, Sung-eui Yoon

The outstanding capability of diffusion models in generating high-quality images poses significant threats when misused by adversaries. In particular, we assume malicious adversaries exploiting diffusion models for inpainting tasks, such as replacing a specific region with a celebrity. While existing methods for protecting images from manipulation in diffusion-based generative models have primarily focused on image-to-image and text-to-image tasks, the challenge of preventing unauthorized inpainting has been rarely addressed, often resulting in suboptimal protection performance. To mitigate inpainting abuses, we propose ADVPAINT, a novel defensive framework that generates adversarial perturbations that effectively disrupt the adversary’s inpainting tasks. ADVPAINT targets the self- and cross-attention blocks in a target diffusion inpainting model to distract semantic understanding and prompt interactions during image generation. ADVPAINT also employs a two-stage perturbation strategy, dividing the perturbation region based on an enlarged bounding box around the object, enhancing robustness across diverse masks of varying shapes and sizes. Our experimental results demonstrate that ADVPAINT’s perturbations are highly effective in disrupting the adversary’s inpainting tasks, outperforming existing methods; ADVPAINT attains over a 100-point increase in FID and substantial decreases in precision.

Applications · Computer Vision

Mor Ventura, Michael Toker, Nitay Calderon, Zorik Gekhman, Yonatan Bitton, Roi Reichart

Will a Visual Language Model (VLM)-based bot warn us about slipping if it detects a wet floor? Recent VLMs have demonstrated impressive capabilities, yet their ability to infer outcomes and causes remains underexplored. To address this, we introduce NL-Eye, a benchmark designed to assess VLMs' visual abductive reasoning skills. NL-Eye adapts the abductive Natural Language Inference (NLI) task to the visual domain, requiring models to evaluate the plausibility of hypothesis images based on a premise image and explain their decisions. NL-Eye consists of 350 carefully curated triplet examples (1,050 images) spanning diverse reasoning categories: physical, functional, logical, emotional, cultural, and social. The data curation process involved two steps—writing textual descriptions and generating images using text-to-image models, both requiring substantial human involvement to ensure high-quality and challenging scenes. Our experiments show that VLMs struggle significantly on NL-Eye, often performing at random baseline levels, while humans excel in both plausibility prediction and explanation quality. This demonstrates a deficiency in the abductive reasoning capabilities of modern VLMs. NL-Eye represents a crucial step toward developing VLMs capable of robust multimodal reasoning for real-world applications, including accident-prevention bots and generated video verification.

Social Aspects · Everything Else

Yue Jiang, Haokun Lin, Yang Bai, Bo Peng, Zhili Liu, Yueming Lyu, Yong Yang, Xingzheng, Jing Dong

Recent studies have discovered that widely used text-to-image diffusion models can replicate training samples during image generation, a phenomenon known as memorization. Existing detection methods primarily focus on identifying memorized prompts. However, in real-world scenarios, image owners may need to verify whether their proprietary or personal images have been memorized by the model, even in the absence of paired prompts or related metadata. We refer to this challenge as image-level memorization detection, where current methods relying on original prompts fall short. In this work, we uncover two characteristics of memorized images after perturbing the inference procedure: lower similarity of the original images and larger magnitudes of TCNP.Building on these insights, we propose Inversion-based Inference Perturbation (IIP), a new framework for image-level memorization detection. Our approach uses unconditional DDIM inversion to derive latent codes that contain core semantic information of original images and optimizes random prompt embeddings to introduce effective perturbation. Memorized images exhibit distinct characteristics within the proposed pipeline, providing a robust basis for detection. To support this task, we construct a comprehensive setup for the image-level memorization detection, carefully curating datasets to simulate realistic memorization scenarios. Using this setup, we evaluate our IIP framework across three different memorization settings, demonstrating its state-of-the-art performance in identifying memorized images in various settings, even in the presence of data augmentation attacks.

Social Aspects · Fairness, Equity, Justice and Safety

Tinghao Xie, Xiangyu Qi, Yi Zeng, Yangsibo Huang, Udari Sehwag, Kaixuan Huang, Luxi He, Boyi Wei, Dacheng Li, Ying Sheng 等

Evaluating aligned large language models' (LLMs) ability to recognize and reject unsafe user requests is crucial for safe, policy-compliant deployments. Existing evaluation efforts, however, face three limitations that we address with **SORRY-Bench**, our proposed benchmark. **First**, existing methods often use coarse-grained taxonomies of unsafe topics, and are over-representing some fine-grained topics. For example, among the ten existing datasets that we evaluated, tests for refusals of self-harm instructions are over 3x less represented than tests for fraudulent activities. SORRY-Bench improves on this by using a fine-grained taxonomy of 44 potentially unsafe topics, and 440 class-balanced unsafe instructions, compiled through human-in-the-loop methods. **Second**, evaluations often overlook the linguistic formatting of prompts, like different languages, dialects, and more --- which are only implicitly considered in many evaluations. We supplement SORRY-bench with 20 diverse linguistic augmentations to systematically examine these effects. **Third**, existing evaluations rely on large LLMs (e.g., GPT-4) for evaluation, which can be computationally expensive. We investigate design choices for creating a fast, accurate automated safety evaluator. By collecting 7K+ human annotations and conducting a meta-evaluation of diverse LLM-as-a-judge designs, we show that fine-tuned 7B LLMs can achieve accuracy comparable to GPT-4 scale LLMs, with lower computational cost. Putting these together, we evaluate over 50 proprietary and open-weight LLMs on SORRY-Bench, analyzing their distinctive safety refusal behaviors. We hope our effort provides a building block for systematic evaluations of LLMs' safety refusal capabilities, in a balanced, granular, and efficient manner. Benchmark demo, data, code, and models are available through [https://sorry-bench.github.io](https://sorry-bench.github.io).

Amrit Poudel, Tim Weninger

Recent decisions to discontinue access to social media APIs are having detrimental effects on Internet research and the field of computational social science as a whole. This lack of access to data has been dubbed the Post-API era of Internet research. Fortunately, popular search engines have the means to crawl, capture, and surface social media data on their Search Engine Results Pages (SERP) if provided the proper search query, and may provide a solution to this dilemma. In the present work we ask: does SERP provide a complete and unbiased sample of social media data? Is SERP a viable alternative to direct API-access? To answer these questions, we perform a comparative analysis between (Google) SERP results and nonsampled data from Reddit and Twitter/X. We find that SERP results are highly biased in favor of popular posts; against political, pornographic, and vulgar posts; are more positive in their sentiment; and have large topical gaps. Overall, we conclude that SERP is not a viable alternative to social media API access.

Salim Chouaki, Abhijnan Chakraborty, Oana Goga, Savvas Zannettou

Understanding how exposure to news on social media impacts public discourse and exacerbates political polarization is a significant endeavor in both computer and social sciences. Unfortunately, progress in this area is hampered by limited access to data due to the closed nature of social media platforms. Consequently, prior studies have been constrained to considering only fragments of users' news exposure and reactions. To overcome this obstacle, we present an innovative measurement approach centered on donating personal data for scientific purposes, facilitated through a privacy-preserving tool that captures users' interactions with news on Facebook. This approach offers a nuanced perspective on users' news exposure and consumption, encompassing different types of news exposure: selective, incidental, algorithmic, and targeted, driven by the diverse underlying mechanisms governing news appearance on users' feeds. Our analysis of data from 472 participants based in the U.S. reveals several interesting findings. For instance, users are more prone to encountering misinformation because of their active selection of low-quality news sources rather than being exposed solely due to friends or platform algorithms. Furthermore, our study uncovers that users are open to engaging with news sources with opposite political ideology as long as these interactions are not visible to their immediate social circles. Overall, our study showcases the viability of data donation as a means to provide clarity to longstanding questions in this field, offering new perspectives on the intricate dynamics of social media news consumption and its effects.

Chen Zhao 0010, Feng Chen 0001, Xintao Wu, Jundong Li, Haifeng Chen

Ethical AI has become increasingly important, and it has been attracting attention from academia and industry, due to its increased popularity in real-world applications with fairness concerns. It also places fundamental importance on ethical considerations in determining legitimate and illegitimate uses of AI. Organizations that apply ethical AI have clearly stated well-defined review processes to ensure adherence to legal guidelines. Therefore, the wave of research at the intersection of ethical AI in data mining and machine learning has also influenced other fields of science, including computer vision, natural language processing, reinforcement learning, and social science. Despite these successes, ethical AI still faces many challenges, such as a lack of interpretable and explainable methods for fairness-aware deep learning models, etc. Consequently, there is an urgent need to bring experts and researchers together at prestigious venues to discuss ethical AI, which has been rarely seen in previous KDD conferences. This workshop will provide a premium platform for both research and industry from different backgrounds to exchange ideas on opportunities, challenges, and cutting-edge techniques in ethical AI.

Emre Eftelioglu, Bistra Dilkina, Naoki Abe, Ramakrishnan Kannan, Yuzhou Chen, Yulia R. Gel, Kathleen Buckingham, Auroop R. Ganguly, James Hodson 0003, Jiafu Mao

The Fragile Earth Workshop is a recurring event in ACM's KDD Conference on research in knowledge discovery and data mining that gathers the research community to find and explore how data science can measure and progress climate and social issues, following the United Nations Sustainable Development Goals (SDGs) framework.

Dunja Mladenic, Dumitru Roman

The European Data Science Day offers a full day focused exclusively on innovative KDD-relevant research and development projects from national and regional funding programs, as well as corporate, start-up, and nonprofit channels. The idea is to bring together a diverse community of researchers in Data Science, Machine Learning, Language Technologies, and Knowledge Discovery, as well as partnerships in the social and physical sciences/arts, to showcase the state-of-the-art in research and applications.

Songgaojun Deng, Maarten de Rijke, Yue Ning 0001

Human events such as hospital visits, protests, and epidemic outbreaks directly affect individuals, communities, and societies. These events are often influenced by factors such as economics, politics, and public policies of our society. The abundance of online data sources such as social networks, official news articles, and personal blogs chronicle societal events, facilitating the development of AI models for social science, public health care, and decision making. Human event modeling generally comprises both the forecasting stage, which estimates future events based on historical data, and interpretation, which seeks to identify influential factors of such events to understand their causative attributes. Recent achievements, fueled by deep learning and the availability of public data, have significantly advanced the field of human event modeling. This survey offers a systematic overview of deep learning technologies for forecasting and interpreting human events, with a primary focus on political events. We first introduce the existing challenges and background in this domain. We then present the problem formulation of event forecasting and interpretation. We investigate recent achievements in graph neural networks, owing to the prevalence of relational data and the efficacy of graph learning models. We also discuss the latest studies that utilize large language models for event reasoning. Lastly, we provide summaries of data resources, open challenges, and future research directions in the study of human event modeling.

Gaku Morio, Soh Young In, Jungah Yoon, Harri Rowlands, Christopher Manning

We introduce ReportParse, a Python-based tool designed to parse corporate sustainability reports. It combines document structure analysis with natural language processing (NLP) models to extract sustainability-related information from the reports. We also provide easy-to-use web and command interfaces. The tool is expected to aid researchers and analysts in evaluating corporate commitment to and management of sustainability efforts.

Pooja Bassin, Abraham G K, Srinath Srinivasa

The localization problem of the United Nations Sustainable Development Goals (SDGs) involves adopting strategies that are in tune with local conditions, to achieve a given SDG target. However, even within a given region, localized conditions may vary drastically. With increasing amounts of Open Government Data (OGD) being available, there is an opportunity to systematically address the localization problem by using predictive and prescriptive modeling techniques. This work presents a predictive and prescriptive modeling dashboard for the SDG indicator maternal deaths (MD) for the Indian state of Karnataka. The dashboard was created by examining a vast set of data points to focus on four factors that showed high correlations with that of MD. We then construct a multivariate linear regression model to showcase the differential impact that a given factor has on the indicator and identify prescribed values for different factors to achieve a given target value of the indicator. Finally, a budget allocation dashboard is also provided that helps policymakers allocate budgets to specific schemes to help operationalize these changes. This dashboard was built by combining data coming from five different OGD sources.

Christopher Morris

Machine learning on graphs, especially using graph neural networks (GNNs), has seen a surge in interest due to the wide availability of graph data across many disciplines, from life and physical to social and engineering sciences. Despite their practical success, our theoretical understanding of the properties of GNNs remains incomplete. Here, we survey the author's and his collaborators' progress in developing a deeper theoretical understanding of GNNs' expressive power and generalization abilities. In addition, we overview recent progress in using GNNs to speed up solvers for hard combinatorial optimization tasks.

Yuhan Liu, Xiuying Chen, Xiaoqing Zhang, Xing Gao, Ji Zhang, Rui Yan

In the digital era, the rapid propagation of fake news and rumors via social networks brings notable societal challenges and impacts public opinion regulation. Traditional fake news modeling typically forecasts the general popularity trends of different groups or numerically represents opinions shift. However, these methods often oversimplify real-world complexities and overlook the rich semantic information of news text. The advent of large language models (LLMs) provides the possibility of modeling subtle dynamics of opinion. Consequently, in this work, we introduce a Fake news Propagation Simulation framework (FPS) based on LLM, which studies the trends and control of fake news propagation in detail. Specifically, each agent in the simulation represents an individual with a distinct personality. They are equipped with both short-term and long-term memory, as well as a reflective mechanism to mimic human-like thinking. Every day, they engage in random opinion exchanges, reflect on their thinking, and update their opinions. Our simulation results uncover patterns in fake news propagation related to topic relevance, and individual traits, aligning with real-world observations. Additionally, we evaluate various intervention strategies and demonstrate that early and appropriately frequent interventions strike a balance between governance cost and effectiveness, offering valuable insights for practical applications. Our study underscores the significant utility and potential of LLMs in combating fake news.

Zihao Wang, Shuyu Li, Tao Zhang, Qi Wang, Pengfei Yu, Jinyang Luo, Yan Liu, Ming Xi, Kejun Zhang

The rapidly evolving multimodal Large Language Models (LLMs) urgently require new benchmarks to uniformly evaluate their performance on understanding and textually describing music. However, due to semantic gaps between Music Information Retrieval (MIR) algorithms and human understanding, discrepancies between professionals and the public, and low precision of annotations, existing music description datasets cannot serve as benchmarks. To this end, we present MuChin, the first open-source music description benchmark in Chinese colloquial language, designed to evaluate the performance of multimodal LLMs in understanding and describing music. We established the Caichong Music Annotation Platform (CaiMAP) that employs an innovative multi-person, multi-stage assurance method, and recruited both amateurs and professionals to ensure the precision of annotations and alignment with popular semantics. Utilizing this method, we built a large-scale, private dataset with multi-dimensional, high-precision music annotations, the Caichong Music Dataset (CaiMD), and carefully selected 1,000 high-quality entries to serve as the test set for MuChin. Based on MuChin, we analyzed the discrepancies between professionals and amateurs in terms of music description, and empirically demonstrated the effectiveness of CaiMD for fine-tuning LLMs. Ultimately, we employed MuChin to evaluate existing music understanding models on their ability to provide colloquial descriptions of music.

Lianwei Wu, Linyong Wang, Yongqiang Zhao

The current approaches for fake news detection are mainly devoted to extracting candidate evidence from comments (or external articles) and establishing interactive reasoning with the news itself to verify the falsehood of the news. However, they still have several drawbacks: 1) The interaction object is coarse-grained, which mainly drives the entire news to participate in interaction, but ignores the learning of potential suspicious segments in news; 2) The reasoning ways are relatively single, making it difficult to explore the various possible correlations between news and candidate evidence. To this end, we propose Unified Evidence Enhancement Inference framework (UEEI) to discover and infer high-quality evidence to reveal the false parts of news for detection. Specifically, UEEI first promotes the interaction fusion between comments and news from the perspectives of semantic and emotion, thereby learning potential suspicious fragments in news. Then, the model constructs entity-level and relationship-level retrievals to screen sufficient candidate evidence from external sources. Finally, we measure coherence between suspicious fragments and candidate evidence by multi-view reasoning, and further infer explainable evidence that discovers the false parts of news. Experiments on three public datasets confirm the effectiveness and interpretability of our UEEI.

Xu Wang, Cheng Li, Yi Chang, Jindong Wang, Yuan Wu

Large Language Models (LLMs) have become integral to a wide spectrum of applications, ranging from traditional computing tasks to advanced artificial intelligence (AI) applications. This widespread adoption has spurred extensive research into LLMs across various disciplines, including the social sciences. Notably, studies have revealed that LLMs possess emotional intelligence, which can be further developed through positive emotional stimuli. This discovery raises an intriguing question: can negative emotions similarly influence LLMs, potentially enhancing their performance? In response to this question, we introduce NegativePrompt, a novel approach underpinned by psychological principles, involving ten specifically designed negative emotional stimuli. We embark on rigorous experimental evaluations of five LLMs including Flan-T5-Large, Vicuna, Llama 2, ChatGPT, and GPT-4, across a set of 45 tasks. The results are revealing: NegativePrompt markedly enhances the performance of LLMs, evidenced by relative improvements of 12.89% in Instruction Induction tasks and 46.25% in BIG-Bench tasks. Moreover, we conduct attention visualization experiments to decipher the underlying mechanisms of NegativePrompt's influence. Our research contributes significantly to the understanding of LLMs and emotion interaction, demonstrating the practical efficacy of NegativePrompt as an emotion-driven method and offering novel insights for the enhancement of LLMs in real-world applications. The code is available at https://github.com/wangxu0820/NegativePrompt.