Search

Large language models can now generate political messages as persuasive as those written by humans, raising concerns about how far this persuasiveness may continue to increase with model size. Here, we generate 720 persuasive messages on 10 U.S. political issues from 24 language models spanning several orders of magnitude in size. We then deploy these messages in a large-scale randomized survey experiment (N = 25,982) to estimate the persuasive capability of each model. Our findings are twofold. First, we find evidence of a log scaling law: model persuasiveness is characterized by sharply diminishing returns, such that current frontier models are barely more persuasive than models smaller in size by an order of magnitude or more. Second, mere task completion (coherence, staying on topic) appears to account for larger models' persuasive advantage. These findings suggest that further scaling model size will not much increase the persuasiveness of static LLM-generated messages., Comment: 16 pages, 4 figures

In this paper, we present the LingOly benchmark, a novel benchmark for advanced reasoning abilities in large language models. Using challenging Linguistic Olympiad puzzles, we evaluate (i) capabilities for in-context identification and generalisation of linguistic patterns in very low-resource or extinct languages, and (ii) abilities to follow complex task instructions. The LingOly benchmark covers more than 90 mostly low-resource languages, minimising issues of data contamination, and contains 1,133 problems across 6 formats and 5 levels of human difficulty. We assess performance with both direct accuracy and comparison to a no-context baseline to penalise memorisation. Scores from 11 state-of-the-art LLMs demonstrate the benchmark to be challenging, and models perform poorly on the higher difficulty problems. On harder problems, even the top model only achieved 38.7% accuracy, 24.7% improvement over the no-context baseline. Large closed models typically outperform open models, and in general, the higher resource the language, the better the scores. These results indicate, in absence of memorisation, true multi-step out-of-domain reasoning remains a challenge for current language models., Comment: 9 pages, 5 figures, 16 pages supplemental materials

Biased information (recently termed bisinformation) continues to be taught in medical curricula, often long after having been debunked. In this paper, we introduce BRICC, a firstin-class initiative that seeks to mitigate medical bisinformation using machine learning to systematically identify and flag text with potential biases, for subsequent review in an expert-in-the-loop fashion, thus greatly accelerating an otherwise labor-intensive process. A gold-standard BRICC dataset was developed throughout several years, and contains over 12K pages of instructional materials. Medical experts meticulously annotated these documents for bias according to comprehensive coding guidelines, emphasizing gender, sex, age, geography, ethnicity, and race. Using this labeled dataset, we trained, validated, and tested medical bias classifiers. We test three classifier approaches: a binary type-specific classifier, a general bias classifier; an ensemble combining bias type-specific classifiers independently-trained; and a multitask learning (MTL) model tasked with predicting both general and type-specific biases. While MTL led to some improvement on race bias detection in terms of F1-score, it did not outperform binary classifiers trained specifically on each task. On general bias detection, the binary classifier achieves up to 0.923 of AUC, a 27.8% improvement over the baseline. This work lays the foundations for debiasing medical curricula by exploring a novel dataset and evaluating different training model strategies. Hence, it offers new pathways for more nuanced and effective mitigation of bisinformation., Comment: Under review

Understanding the writing frame of news articles is vital for addressing social issues, and thus has attracted notable attention in the fields of communication studies. Yet, assessing such news article frames remains a challenge due to the absence of a concrete and unified standard dataset that considers the comprehensive nuances within news content. To address this gap, we introduce an extended version of a large labeled news article dataset with 16,687 new labeled pairs. Leveraging the pairwise comparison of news articles, our method frees the work of manual identification of frame classes in traditional news frame analysis studies. Overall we introduce the most extensive cross-lingual news article similarity dataset available to date with 26,555 labeled news article pairs across 10 languages. Each data point has been meticulously annotated according to a codebook detailing eight critical aspects of news content, under a human-in-the-loop framework. Application examples demonstrate its potential in unearthing country communities within global news coverage, exposing media bias among news outlets, and quantifying the factors related to news creation. We envision that this news similarity dataset will broaden our understanding of the media ecosystem in terms of news coverage of events and perspectives across countries, locations, languages, and other social constructs. By doing so, it can catalyze advancements in social science research and applied methodologies, thereby exerting a profound impact on our society.

Diaspora communities are disproportionately impacted by off-the-radar misinformation and often neglected by mainstream fact-checking efforts, creating a critical need to scale-up efforts of nascent fact-checking initiatives. In this paper we present SynDy, a framework for Synthetic Dynamic Dataset Generation to leverage the capabilities of the largest frontier Large Language Models (LLMs) to train local, specialized language models. To the best of our knowledge, SynDy is the first paper utilizing LLMs to create fine-grained synthetic labels for tasks of direct relevance to misinformation mitigation, namely Claim Matching, Topical Clustering, and Claim Relationship Classification. SynDy utilizes LLMs and social media queries to automatically generate distantly-supervised, topically-focused datasets with synthetic labels on these three tasks, providing essential tools to scale up human-led fact-checking at a fraction of the cost of human-annotated data. Training on SynDy's generated labels shows improvement over a standard baseline and is not significantly worse compared to training on human labels (which may be infeasible to acquire). SynDy is being integrated into Meedan's chatbot tiplines that are used by over 50 organizations, serve over 230K users annually, and automatically distribute human-written fact-checks via messaging apps such as WhatsApp. SynDy will also be integrated into our deployed Co-Insights toolkit, enabling low-resource organizations to launch tiplines for their communities. Finally, we envision SynDy enabling additional fact-checking tools such as matching new misinformation claims to high-quality explainers on common misinformation topics.

News coverage profoundly affects how countries and individuals behave in international relations. Yet, we have little empirical evidence of how news coverage varies across countries. To enable studies of global news coverage, we develop an efficient computational methodology that comprises three components: (i) a transformer model to estimate multilingual news similarity; (ii) a global event identification system that clusters news based on a similarity network of news articles; and (iii) measures of news synchrony across countries and news diversity within a country, based on country-specific distributions of news coverage of the global events. Each component achieves state-of-the art performance, scaling seamlessly to massive datasets of millions of news articles. We apply the methodology to 60 million news articles published globally between January 1 and June 30, 2020, across 124 countries and 10 languages, detecting 4357 news events. We identify the factors explaining diversity and synchrony of news coverage across countries. Our study reveals that news media tend to cover a more diverse set of events in countries with larger Internet penetration, more official languages, larger religious diversity, higher economic inequality, and larger populations. Coverage of news events is more synchronized between countries that not only actively participate in commercial and political relations -- such as, pairs of countries with high bilateral trade volume, and countries that belong to the NATO military alliance or BRICS group of major emerging economies -- but also countries that share certain traits: an official language, high GDP, and high democracy indices.

Perceptions of hate can vary greatly across cultural contexts. Hate speech (HS) datasets, however, have traditionally been developed by language. This hides potential cultural biases, as one language may be spoken in different countries home to different cultures. In this work, we evaluate cultural bias in HS datasets by leveraging two interrelated cultural proxies: language and geography. We conduct a systematic survey of HS datasets in eight languages and confirm past findings on their English-language bias, but also show that this bias has been steadily decreasing in the past few years. For three geographically-widespread languages -- English, Arabic and Spanish -- we then leverage geographical metadata from tweets to approximate geo-cultural contexts by pairing language and country information. We find that HS datasets for these languages exhibit a strong geo-cultural bias, largely overrepresenting a handful of countries (e.g., US and UK for English) relative to their prominence in both the broader social media population and the general population speaking these languages. Based on these findings, we formulate recommendations for the creation of future HS datasets., Comment: Accepted at WOAH (NAACL 2024)

Human feedback plays a central role in the alignment of Large Language Models (LLMs). However, open questions remain about the methods (how), domains (where), people (who) and objectives (to what end) of human feedback collection. To navigate these questions, we introduce PRISM, a new dataset which maps the sociodemographics and stated preferences of 1,500 diverse participants from 75 countries, to their contextual preferences and fine-grained feedback in 8,011 live conversations with 21 LLMs. PRISM contributes (i) wide geographic and demographic participation in human feedback data; (ii) two census-representative samples for understanding collective welfare (UK and US); and (iii) individualised feedback where every rating is linked to a detailed participant profile, thus permitting exploration of personalisation and attribution of sample artefacts. We focus on collecting conversations that centre subjective and multicultural perspectives on value-laden and controversial topics, where we expect the most interpersonal and cross-cultural disagreement. We demonstrate the usefulness of PRISM via three case studies of dialogue diversity, preference diversity, and welfare outcomes, showing that it matters which humans set alignment norms. As well as offering a rich community resource, we advocate for broader participation in AI development and a more inclusive approach to technology design.

This paper introduces v0.5 of the AI Safety Benchmark, which has been created by the MLCommons AI Safety Working Group. The AI Safety Benchmark has been designed to assess the safety risks of AI systems that use chat-tuned language models. We introduce a principled approach to specifying and constructing the benchmark, which for v0.5 covers only a single use case (an adult chatting to a general-purpose assistant in English), and a limited set of personas (i.e., typical users, malicious users, and vulnerable users). We created a new taxonomy of 13 hazard categories, of which 7 have tests in the v0.5 benchmark. We plan to release version 1.0 of the AI Safety Benchmark by the end of 2024. The v1.0 benchmark will provide meaningful insights into the safety of AI systems. However, the v0.5 benchmark should not be used to assess the safety of AI systems. We have sought to fully document the limitations, flaws, and challenges of v0.5. This release of v0.5 of the AI Safety Benchmark includes (1) a principled approach to specifying and constructing the benchmark, which comprises use cases, types of systems under test (SUTs), language and context, personas, tests, and test items; (2) a taxonomy of 13 hazard categories with definitions and subcategories; (3) tests for seven of the hazard categories, each comprising a unique set of test items, i.e., prompts. There are 43,090 test items in total, which we created with templates; (4) a grading system for AI systems against the benchmark; (5) an openly available platform, and downloadable tool, called ModelBench that can be used to evaluate the safety of AI systems on the benchmark; (6) an example evaluation report which benchmarks the performance of over a dozen openly available chat-tuned language models; (7) a test specification for the benchmark.

Gun violence is a pressing and growing human rights issue that affects nearly every dimension of the social fabric, from healthcare and education to psychology and the economy. Reliable data on firearm events is paramount to developing more effective public policy and emergency responses. However, the lack of comprehensive databases and the risks of in-person surveys prevent human rights organizations from collecting needed data in most countries. Here, we partner with a Brazilian human rights organization to conduct a systematic evaluation of language models to assist with monitoring real-world firearm events from social media data. We propose a fine-tuned BERT-based model trained on Twitter (now X) texts to distinguish gun violence reports from ordinary Portuguese texts. Our model achieves a high AUC score of 0.97. We then incorporate our model into a web application and test it in a live intervention. We study and interview Brazilian analysts who continuously fact-check social media texts to identify new gun violence events. Qualitative assessments show that our solution helped all analysts use their time more efficiently and expanded their search capacities. Quantitative assessments show that the use of our model was associated with more analysts' interactions with online users reporting gun violence. Taken together, our findings suggest that modern Natural Language Processing techniques can help support the work of human rights organizations.

This study analyzes misinformation from WhatsApp, Twitter, and Kwai during the 2022 Brazilian general election. Given the democratic importance of accurate information during elections, multiple fact-checking organizations collaborated to identify and respond to misinformation via WhatsApp tiplines and power a fact-checking feature within a chatbot operated by Brazil's election authority, the TSE. WhatsApp is installed on over 99% of smartphones in Brazil, and the TSE chatbot was used by millions of citizens in the run-up to the elections. During the same period, we collected social media data from Twitter (now X) and Kwai (a popular video-sharing app similar to TikTok). Using the WhatsApp, Kwai, and Twitter data along with fact-checks from three Brazilian fact-checking organizations, we find unique claims on each platform. Even when the same claims are present on different platforms, they often differ in format, detail, length, or other characteristics. Our research highlights the limitations of current claim matching algorithms to match claims across platforms with such differences and identifies areas for further algorithmic development. Finally, we perform a descriptive analysis examining the formats (image, video, audio, text) and content themes of popular misinformation claims.

The past year has seen rapid acceleration in the development of large language models (LLMs). However, without proper steering and safeguards, LLMs will readily follow malicious instructions, provide unsafe advice, and generate toxic content. We introduce SimpleSafetyTests (SST) as a new test suite for rapidly and systematically identifying such critical safety risks. The test suite comprises 100 test prompts across five harm areas that LLMs, for the vast majority of applications, should refuse to comply with. We test 11 open-access and open-source LLMs and four closed-source LLMs, and find critical safety weaknesses. While some of the models do not give a single unsafe response, most give unsafe responses to more than 20% of the prompts, with over 50% unsafe responses in the extreme. Prepending a safety-emphasising system prompt substantially reduces the occurrence of unsafe responses, but does not completely stop them from happening. Trained annotators labelled every model response to SST (n = 3,000). We use these annotations to evaluate five AI safety filters (which assess whether a models' response is unsafe given a prompt) as a way of automatically evaluating models' performance on SST. The filters' performance varies considerably. There are also differences across the five harm areas, and on the unsafe versus safe responses. The widely-used Perspective API has 72% accuracy and a newly-created zero-shot prompt to OpenAI's GPT-4 performs best with 89% accuracy. Content Warning: This paper contains prompts and responses that relate to child abuse, suicide, self-harm and eating disorders, scams and fraud, illegal items, and physical harm.

Misinformation and disinformation are growing threats in the digital age, spreading rapidly across languages and borders. This paper investigates the prevalence and dynamics of multilingual misinformation through an analysis of over 250,000 unique fact-checks spanning 95 languages. First, we find that while the majority of misinformation claims are only fact-checked once, 11.7%, corresponding to more than 21,000 claims, are checked multiple times. Using fact-checks as a proxy for the spread of misinformation, we find 33% of repeated claims cross linguistic boundaries, suggesting that some misinformation permeates language barriers. However, spreading patterns exhibit strong homophily, with misinformation more likely to spread within the same language. To study the evolution of claims over time and mutations across languages, we represent fact-checks with multilingual sentence embeddings and cluster semantically similar claims. We analyze the connected components and shortest paths connecting different versions of a claim finding that claims gradually drift over time and undergo greater alteration when traversing languages. Overall, this novel investigation of multilingual misinformation provides key insights. It quantifies redundant fact-checking efforts, establishes that some claims diffuse across languages, measures linguistic homophily, and models the temporal and cross-lingual evolution of claims. The findings advocate for expanded information sharing between fact-checkers globally while underscoring the importance of localized verification.

Human feedback is increasingly used to steer the behaviours of Large Language Models (LLMs). However, it is unclear how to collect and incorporate feedback in a way that is efficient, effective and unbiased, especially for highly subjective human preferences and values. In this paper, we survey existing approaches for learning from human feedback, drawing on 95 papers primarily from the ACL and arXiv repositories.First, we summarise the past, pre-LLM trends for integrating human feedback into language models. Second, we give an overview of present techniques and practices, as well as the motivations for using feedback; conceptual frameworks for defining values and preferences; and how feedback is collected and from whom. Finally, we encourage a better future of feedback learning in LLMs by raising five unresolved conceptual and practical challenges., Comment: Accepted for the 2023 Conference on Empirical Methods in Natural Language Processing (EMNLP, Main)

The emergence of tools based on Large Language Models (LLMs), such as OpenAI's ChatGPT, Microsoft's Bing Chat, and Google's Bard, has garnered immense public attention. These incredibly useful, natural-sounding tools mark significant advances in natural language generation, yet they exhibit a propensity to generate false, erroneous, or misleading content -- commonly referred to as "hallucinations." Moreover, LLMs can be exploited for malicious applications, such as generating false but credible-sounding content and profiles at scale. This poses a significant challenge to society in terms of the potential deception of users and the increasing dissemination of inaccurate information. In light of these risks, we explore the kinds of technological innovations, regulatory reforms, and AI literacy initiatives needed from fact-checkers, news organizations, and the broader research and policy communities. By identifying the risks, the imminent threats, and some viable solutions, we seek to shed light on navigating various aspects of veracity in the era of generative AI., Comment: Our article offers a comprehensive examination of the challenges and risks associated with Large Language Models (LLMs), focusing on their potential impact on the veracity of information in today's digital landscape

In this paper, we address the concept of "alignment" in large language models (LLMs) through the lens of post-structuralist socio-political theory, specifically examining its parallels to empty signifiers. To establish a shared vocabulary around how abstract concepts of alignment are operationalised in empirical datasets, we propose a framework that demarcates: 1) which dimensions of model behaviour are considered important, then 2) how meanings and definitions are ascribed to these dimensions, and by whom. We situate existing empirical literature and provide guidance on deciding which paradigm to follow. Through this framework, we aim to foster a culture of transparency and critical evaluation, aiding the community in navigating the complexities of aligning LLMs with human populations., Comment: Socially Responsible Language Modelling Research (SoLaR) @ NeurIPs 2023

Large Language Models (LLMs), now used daily by millions, can encode societal biases, exposing their users to representational harms. A large body of scholarship on LLM bias exists but it predominantly adopts a Western-centric frame and attends comparatively less to bias levels and potential harms in the Global South. In this paper, we quantify stereotypical bias in popular LLMs according to an Indian-centric frame through Indian-BhED, a first of its kind dataset, containing stereotypical and anti-stereotypical examples in the context of caste and religious stereotypes in India. We find that the majority of LLMs tested have a strong propensity to output stereotypes in the Indian context, especially when compared to axes of bias traditionally studied in the Western context, such as gender and race. Notably, we find that GPT-2, GPT-2 Large, and GPT 3.5 have a particularly high propensity for preferring stereotypical outputs as a percent of all sentences for the axes of caste (63-79%) and religion (69-72%). We finally investigate potential causes for such harmful behaviour in LLMs, and posit intervention techniques to reduce both stereotypical and anti-stereotypical biases. The findings of this work highlight the need for including more diverse voices when researching fairness in AI and evaluating LLMs., Comment: To be published in GoodIT '24, doi:10.1145/3677525.3678666. 14 pages

Public figures receive a disproportionate amount of abuse on social media, impacting their active participation in public life. Automated systems can identify abuse at scale but labelling training data is expensive, complex and potentially harmful. So, it is desirable that systems are efficient and generalisable, handling both shared and specific aspects of online abuse. We explore the dynamics of cross-group text classification in order to understand how well classifiers trained on one domain or demographic can transfer to others, with a view to building more generalisable abuse classifiers. We fine-tune language models to classify tweets targeted at public figures across DOmains (sport and politics) and DemOgraphics (women and men) using our novel DODO dataset, containing 28,000 labelled entries, split equally across four domain-demographic pairs. We find that (i) small amounts of diverse data are hugely beneficial to generalisation and model adaptation; (ii) models transfer more easily across demographics but models trained on cross-domain data are more generalisable; (iii) some groups contribute more to generalisability than others; and (iv) dataset similarity is a signal of transferability., Comment: 15 pages, 7 figures, 4 tables

Large language models (LLMs) are used to generate content for a wide range of tasks, and are set to reach a growing audience in coming years due to integration in product interfaces like ChatGPT or search engines like Bing. This intensifies the need to ensure that models are aligned with human preferences and do not produce unsafe, inaccurate or toxic outputs. While alignment techniques like reinforcement learning with human feedback (RLHF) and red-teaming can mitigate some safety concerns and improve model capabilities, it is unlikely that an aggregate fine-tuning process can adequately represent the full range of users' preferences and values. Different people may legitimately disagree on their preferences for language and conversational norms, as well as on values or ideologies which guide their communication. Personalising LLMs through micro-level preference learning processes may result in models that are better aligned with each user. However, there are several normative challenges in defining the bounds of a societally-acceptable and safe degree of personalisation. In this paper, we ask how, and in what ways, LLMs should be personalised. First, we review literature on current paradigms for aligning LLMs with human feedback, and identify issues including (i) a lack of clarity regarding what alignment means; (ii) a tendency of technology providers to prescribe definitions of inherently subjective preferences and values; and (iii) a 'tyranny of the crowdworker', exacerbated by a lack of documentation in who we are really aligning to. Second, we present a taxonomy of benefits and risks associated with personalised LLMs, for individuals and society at large. Finally, we propose a three-tiered policy framework that allows users to experience the benefits of personalised alignment, while restraining unsafe and undesirable LLM-behaviours within (supra-)national and organisational bounds., Comment: 19 pages, 1 table

The spread of misinformation is a pressing global problem that has elicited a range of responses from researchers, policymakers, civil society and industry. Over the past decade, these stakeholders have developed many interventions to tackle misinformation that vary across factors such as which effects of misinformation they hope to target, at what stage in the misinformation lifecycle they are aimed at, and who they are implemented by. These interventions also differ in how effective they are at reducing susceptibility to (and curbing the spread of) misinformation. In recent years, a vast amount of scholarly work on misinformation has become available, which extends across multiple disciplines and methodologies. It has become increasingly difficult to comprehensively map all of the available interventions, assess their efficacy, and understand the challenges, opportunities and tradeoffs associated with using them. Few papers have systematically assessed and compared the various interventions, which has led to a lack of understanding in civic and policymaking discourses. With this in mind, we develop a new hierarchical framework for understanding interventions against misinformation online. The framework comprises three key elements: Interventions that Prepare people to be less susceptible; Interventions that Curb the spread and effects of misinformation; and Interventions that Respond to misinformation. We outline how different interventions are thought to work, categorise them, and summarise the available evidence on their efficacy; offering researchers, policymakers and practitioners working to combat online misinformation both an analytical framework that they can use to understand and evaluate different interventions (and which could be extended to address new interventions that we do not describe here) and a summary of the range of interventions that have been proposed to date.

We propose a novel system to help fact-checkers formulate search queries for known misinformation claims and effectively search across multiple social media platforms. We introduce an adaptable rewriting strategy, where editing actions for queries containing claims (e.g., swap a word with its synonym; change verb tense into present simple) are automatically learned through offline reinforcement learning. Our model uses a decision transformer to learn a sequence of editing actions that maximizes query retrieval metrics such as mean average precision. We conduct a series of experiments showing that our query rewriting system achieves a relative increase in the effectiveness of the queries of up to 42%, while producing editing action sequences that are human interpretable., Comment: AACL 2023 (long paper)

Annotating abusive language is expensive, logistically complex and creates a risk of psychological harm. However, most machine learning research has prioritized maximizing effectiveness (i.e., F1 or accuracy score) rather than data efficiency (i.e., minimizing the amount of data that is annotated). In this paper, we use simulated experiments over two datasets at varying percentages of abuse to demonstrate that transformers-based active learning is a promising approach to substantially raise efficiency whilst still maintaining high effectiveness, especially when abusive content is a smaller percentage of the dataset. This approach requires a fraction of labeled data to reach performance equivalent to training over the full dataset., Comment: Third Workshop on Threat, Aggression and Cyberbullying (COLING 2022)

Polarization and echo chambers are often studied in the context of explicitly political events such as elections, and little scholarship has examined the mixing of political groups in non-political contexts. A major obstacle to studying political polarization in non-political contexts is that political leaning (i.e., left vs right orientation) is often unknown. Nonetheless, political leaning is known to correlate (sometimes quite strongly) with many lifestyle choices leading to stereotypes such as the "latte-drinking liberal." We develop a machine learning classifier to infer political leaning from non-political text and, optionally, the accounts a user follows on social media. We use Voter Advice Application results shared on Twitter as our groundtruth and train and test our classifier on a Twitter dataset comprising the 3,200 most recent tweets of each user after removing any tweets with political text. We correctly classify the political leaning of most users (F1 scores range from 0.70 to 0.85 depending on coverage). We find no relationship between the level of political activity and our classification results. We apply our classifier to a case study of news sharing in the UK and discover that, in general, the sharing of political news exhibits a distinctive left-right divide while sports news does not.

An important challenge for news fact-checking is the effective dissemination of existing fact-checks. This in turn brings the need for reliable methods to detect previously fact-checked claims. In this paper, we focus on automatically finding existing fact-checks for claims made in social media posts (tweets). We conduct both classification and retrieval experiments, in monolingual (English only), multilingual (Spanish, Portuguese), and cross-lingual (Hindi-English) settings using multilingual transformer models such as XLM-RoBERTa and multilingual embeddings such as LaBSE and SBERT. We present promising results for "match" classification (86% average accuracy) in four language pairs. We also find that a BM25 baseline outperforms or is on par with state-of-the-art multilingual embedding models for the retrieval task during our monolingual experiments. We highlight and discuss NLP challenges while addressing this problem in different languages, and we introduce a novel curated dataset of fact-checks and corresponding tweets for future research., Comment: Accepted to De-Factify Workshop at AAAI 2022

Most Fairness in AI research focuses on exposing biases in AI systems. A broader lens on fairness reveals that AI can serve a greater aspiration: rooting out societal inequities from their source. Specifically, we focus on inequities in health information, and aim to reduce bias in that domain using AI. The AI algorithms under the hood of search engines and social media, many of which are based on recommender systems, have an outsized impact on the quality of medical and health information online. Therefore, embedding bias detection and reduction into these recommender systems serving up medical and health content online could have an outsized positive impact on patient outcomes and wellbeing. In this position paper, we offer the following contributions: (1) we propose a novel framework of Fairness via AI, inspired by insights from medical education, sociology and antiracism; (2) we define a new term, bisinformation, which is related to, but distinct from, misinformation, and encourage researchers to study it; (3) we propose using AI to study, detect and mitigate biased, harmful, and/or false health information that disproportionately hurts minority groups in society; and (4) we suggest several pillars and pose several open problems in order to seed inquiry in this new space. While part (3) of this work specifically focuses on the health domain, the fundamental computer science advances and contributions stemming from research efforts in bias reduction and Fairness via AI have broad implications in all areas of society., Comment: To appear in: The 4th FAccTRec Workshop on Responsible Recommendation at RecSys 2021

Detecting online hate is a complex task, and low-performing models have harmful consequences when used for sensitive applications such as content moderation. Emoji-based hate is an emerging challenge for automated detection. We present HatemojiCheck, a test suite of 3,930 short-form statements that allows us to evaluate performance on hateful language expressed with emoji. Using the test suite, we expose weaknesses in existing hate detection models. To address these weaknesses, we create the HatemojiBuild dataset using a human-and-model-in-the-loop approach. Models built with these 5,912 adversarial examples perform substantially better at detecting emoji-based hate, while retaining strong performance on text-only hate. Both HatemojiCheck and HatemojiBuild are made publicly available. See our Github Repository (https://github.com/HannahKirk/Hatemoji). HatemojiCheck, HatemojiBuild, and the final Hatemoji Model are also available on HuggingFace (https://huggingface.co/datasets/HannahRoseKirk/).

Accurate detection and classification of online hate is a difficult task. Implicit hate is particularly challenging as such content tends to have unusual syntax, polysemic words, and fewer markers of prejudice (e.g., slurs). This problem is heightened with multimodal content, such as memes (combinations of text and images), as they are often harder to decipher than unimodal content (e.g., text alone). This paper evaluates the role of semantic and multimodal context for detecting implicit and explicit hate. We show that both text- and visual- enrichment improves model performance, with the multimodal model (0.771) outperforming other models' F1 scores (0.544, 0.737, and 0.754). While the unimodal-text context-aware (transformer) model was the most accurate on the subtask of implicit hate detection, the multimodal model outperformed it overall because of a lower propensity towards false positives. We find that all models perform better on content with full annotator agreement and that multimodal models are best at classifying the content where annotators disagree. To conduct these investigations, we undertook high-quality annotation of a sample of 5,000 multimodal entries. Tweets were annotated for primary category, modality, and strategy. We make this corpus, along with the codebook, code, and final model, freely available., Comment: Please note the paper contains examples of hateful content

There is currently no easy way to fact-check content on WhatsApp and other end-to-end encrypted platforms at scale. In this paper, we analyze the usefulness of a crowd-sourced "tipline" through which users can submit content ("tips") that they want fact-checked. We compare the tips sent to a WhatsApp tipline run during the 2019 Indian national elections with the messages circulating in large, public groups on WhatsApp and other social media platforms during the same period. We find that tiplines are a very useful lens into WhatsApp conversations: a significant fraction of messages and images sent to the tipline match with the content being shared on public WhatsApp groups and other social media. Our analysis also shows that tiplines cover the most popular content well, and a majority of such content is often shared to the tipline before appearing in large, public WhatsApp groups. Overall, our findings suggest tiplines can be an effective source for discovering content to fact-check.

Manual fact-checking does not scale well to serve the needs of the internet. This issue is further compounded in non-English contexts. In this paper, we discuss claim matching as a possible solution to scale fact-checking. We define claim matching as the task of identifying pairs of textual messages containing claims that can be served with one fact-check. We construct a novel dataset of WhatsApp tipline and public group messages alongside fact-checked claims that are first annotated for containing "claim-like statements" and then matched with potentially similar items and annotated for claim matching. Our dataset contains content in high-resource (English, Hindi) and lower-resource (Bengali, Malayalam, Tamil) languages. We train our own embedding model using knowledge distillation and a high-quality "teacher" model in order to address the imbalance in embedding quality between the low- and high-resource languages in our dataset. We provide evaluations on the performance of our solution and compare with baselines and existing state-of-the-art multilingual embedding models, namely LASER and LaBSE. We demonstrate that our performance exceeds LASER and LaBSE in all settings. We release our annotated datasets, codebooks, and trained embedding model to allow for further research., Comment: to appear in ACL 2021 as a long paper

The semantics of emoji has, to date, been considered from a static perspective. We offer the first longitudinal study of how emoji semantics changes over time, applying techniques from computational linguistics to six years of Twitter data. We identify five patterns in emoji semantic development and find evidence that the less abstract an emoji is, the more likely it is to undergo semantic change. In addition, we analyse select emoji in more detail, examining the effect of seasonality and world events on emoji semantics. To aid future work on emoji and semantics, we make our data publicly available along with a web-based interface that anyone can use to explore semantic change in emoji.

Online hate is a growing concern on many social media platforms and other sites. To combat it, technology companies are increasingly identifying and sanctioning `hateful users' rather than simply moderating hateful content. Yet, most research in online hate detection to date has focused on hateful content. This paper examines how fairer and more accurate hateful user detection systems can be developed by incorporating social network information through geometric deep learning. Geometric deep learning dynamically learns information-rich network representations and can generalise to unseen nodes. This is essential for moving beyond manually engineered network features, which lack scalability and produce information-sparse network representations. This paper compares the accuracy of geometric deep learning with other techniques which either exclude network information or incorporate it through manual feature engineering (e.g., node2vec). It also evaluates the fairness of these techniques using the `predictive equality' criteria, comparing the false positive rates on a subset of 136 African-American users with 4836 other users. Geometric deep learning produces the most accurate and fairest classifier, with an AUC score of 90.8\% on the entire dataset and a false positive rate of zero among the African-American subset for the best performing model. This highlights the benefits of more effectively incorporating social network features in automated hateful user detection. Such an approach is also easily operationalized for real-world content moderation as it has an efficient and scalable design.

Online government petitions represent a new data-rich mode of political participation. This work examines the thus far understudied dynamics of sharing petitions on social media in order to garner signatures and, ultimately, a government response. Using 20 months of Twitter data comprising over 1 million tweets linking to a petition, we perform analyses of networks constructed of petitions and supporters on Twitter, revealing implicit social dynamics therein. We find that Twitter users do not exclusively share petitions on one issue nor do they share exclusively popular petitions. Among the over 240,000 Twitter users, we find latent support groups, with the most central users primarily being politically active "average" individuals. Twitter as a platform for sharing government petitions, thus, appears to hold potential to foster the creation of and coordination among a new form of latent support interest groups online., Comment: Presented at IPP2016 Conference, Oxford, UK. http://blogs.oii.ox.ac.uk/ipp-conference/2016.html

The outbreak of COVID-19 has transformed societies across the world as governments tackle the health, economic and social costs of the pandemic. It has also raised concerns about the spread of hateful language and prejudice online, especially hostility directed against East Asia. In this paper we report on the creation of a classifier that detects and categorizes social media posts from Twitter into four classes: Hostility against East Asia, Criticism of East Asia, Meta-discussions of East Asian prejudice and a neutral class. The classifier achieves an F1 score of 0.83 across all four classes. We provide our final model (coded in Python), as well as a new 20,000 tweet training dataset used to make the classifier, two analyses of hashtags associated with East Asian prejudice and the annotation codebook. The classifier can be implemented by other researchers, assisting with both online content moderation processes and further research into the dynamics, prevalence and impact of East Asian prejudice online during this global pandemic., Comment: 12 pages

Social media has become an emerging alternative to opinion polls for public opinion collection, while it is still posing many challenges as a passive data source, such as structurelessness, quantifiability, and representativeness. Social media data with geotags provide new opportunities to unveil the geographic locations of users expressing their opinions. This paper aims to answer two questions: 1) whether quantifiable measurement of public opinion can be obtained from social media and 2) whether it can produce better or complementary measures compared to opinion polls. This research proposes a novel approach to measure the relative opinion of Twitter users towards public issues in order to accommodate more complex opinion structures and take advantage of the geography pertaining to the public issues. To ensure that this new measure is technically feasible, a modeling framework is developed including building a training dataset by adopting a state-of-the-art approach and devising a new deep learning method called Opinion-Oriented Word Embedding. With a case study of the tweets selected for the 2016 U.S. presidential election, we demonstrate the predictive superiority of our relative opinion approach and we show how it can aid visual analytics and support opinion predictions. Although the relative opinion measure is proved to be more robust compared to polling, our study also suggests that the former can advantageously complement the later in opinion prediction.

Accurate understanding and forecasting of traffic is a key contemporary problem for policymakers. Road networks are increasingly congested, yet traffic data is often expensive to obtain, making informed policy-making harder. This paper explores the extent to which traffic disruption can be estimated from static features from the volunteered geographic information site OpenStreetMap (OSM). We use OSM features as predictors for linear regressions of counts of traffic disruptions and traffic volume at 6,500 points in the road network within 112 regions of Oxfordshire, UK. We show that more than half the variation in traffic volume and disruptions can be explained with static features alone, and use cross-validation and recursive feature elimination to evaluate the predictive power and importance of different land use categories. Finally, we show that using OSM's granular point of interest data allows for better predictions than the aggregate categories typically used in studies of transportation and land use., Comment: 2 figures, 5 tables

Accurate modelling of local population movement patterns is a core contemporary concern for urban policymakers, affecting both the short term deployment of public transport resources and the longer term planning of transport infrastructure. Yet, while macro-level population movement models (such as the gravity and radiation models) are well developed, micro-level alternatives are in much shorter supply, with most macro-models known to perform badly in smaller geographic confines. In this paper we take a first step to remedying this deficit, by leveraging two novel datasets to analyse where and why macro-level models of human mobility break down at small scales. In particular, we use an anonymised aggregate dataset from a major mobility app and combine this with freely available data from OpenStreetMap concerning land-use composition of different areas around the county of Oxfordshire in the United Kingdom. We show where different models fail, and make the case for a new modelling strategy which moves beyond rough heuristics such as distance and population size towards a detailed, granular understanding of the opportunities presented in different areas of the city.

Social media provide access to behavioural data at an unprecedented scale and granularity. However, using these data to understand phenomena in a broader population is difficult due to their non-representativeness and the bias of statistical inference tools towards dominant languages and groups. While demographic attribute inference could be used to mitigate such bias, current techniques are almost entirely monolingual and fail to work in a global environment. We address these challenges by combining multilingual demographic inference with post-stratification to create a more representative population sample. To learn demographic attributes, we create a new multimodal deep neural architecture for joint classification of age, gender, and organization-status of social media users that operates in 32 languages. This method substantially outperforms current state of the art while also reducing algorithmic bias. To correct for sampling biases, we propose fully interpretable multilevel regression methods that estimate inclusion probabilities from inferred joint population counts and ground-truth population counts. In a large experiment over multilingual heterogeneous European regions, we show that our demographic inference and bias correction together allow for more accurate estimates of populations and make a significant step towards representative social sensing in downstream applications with multilingual social media., Comment: 12 pages, 10 figures, Proceedings of the 2019 World Wide Web Conference (WWW '19)

Recent election surprises, regime changes, and political shocks indicate that political agendas have become more fast-moving and volatile. The ability to measure the complex dynamics of agenda change and capture the nature and extent of volatility in political systems is therefore more crucial than ever before. This study proposes a definition and operationalization of volatility that combines insights from political science, communications, information theory, and computational techniques. The proposed measures of fractionalization and agenda change encompass the shifting salience of issues in the agenda as a whole and allow the study of agendas across different domains. We evaluate these metrics and compare them to other measures such as issue-level survival rates and the Pedersen Index, which uses public-opinion poll data to measure public agendas, as well as traditional media content to measure media agendas in the UK and Germany. We show how these measures complement existing approaches and could be employed in future agenda-setting research., Comment: Copyright is held by the authors, and the article is published by Oxford University Press

The Internet has become a fundamental resource for activism as it facilitates political mobilization at a global scale. Petition platforms are a clear example of how thousands of people around the world can contribute to social change. Avaaz.org, with a presence in over 200 countries, is one of the most popular of this type. However, little research has focused on this platform, probably due to a lack of available data. In this work we retrieved more than 350K petitions, standardized their field values, and added new information using language detection and named-entity recognition. To motivate future research with this unique repository of global protest, we present a first exploration of the dataset. In particular, we examine how social media campaigning is related to the success of petitions, as well as some geographic and linguistic findings about the worldwide community of Avaaz.org. We conclude with example research questions that could be addressed with our dataset., Comment: Accepted as a dataset paper at the 12th International AAAI Conference on Web and Social Media (ICWSM-18). This preprint includes an additional appendix with the reasons, provided by Avaaz.org, about the anomalies detected when exploring the dataset. For academic purposes, please cite the ICWSM version

Social media are now a routine part of political campaigns all over the world. However, studies of the impact of campaigning on social platform have thus far been limited to cross-sectional datasets from one election period which are vulnerable to unobserved variable bias. Hence empirical evidence on the effectiveness of political social media activity is thin. We address this deficit by analysing a novel panel dataset of political Twitter activity in the 2015 and 2017 elections in the United Kingdom. We find that Twitter based campaigning does seem to help win votes, a finding which is consistent across a variety of different model specifications including a first difference regression. The impact of Twitter use is small in absolute terms, though comparable with that of campaign spending. Our data also support the idea that effects are mediated through other communication channels, hence challenging the relevance of engaging in an interactive fashion.

The number and quality of user reviews greatly affects consumer purchasing decisions. While reviews in all languages are increasing, it is still often the case (especially for non-English speakers) that there are only a few reviews in a person's first language. Using an online experiment, we examine the value that potential purchasers receive from interfaces showing additional reviews in a second language. The results paint a complicated picture with both positive and negative reactions to the inclusion of foreign-language reviews. Roughly 26-28% of subjects clicked to see translations of the foreign-language content when given the opportunity, and those who did so were more likely to select the product with foreign-language reviews than those who did not.

The emergence of large stores of transactional data generated by increasing use of digital devices presents a huge opportunity for policymakers to improve their knowledge of the local environment and thus make more informed and better decisions. A research frontier is hence emerging which involves exploring the type of measures that can be drawn from data stores such as mobile phone logs, Internet searches and contributions to social media platforms, and the extent to which these measures are accurate reflections of the wider population. This paper contributes to this research frontier, by exploring the extent to which local commuting patterns can be estimated from data drawn from Twitter. It makes three contributions in particular. First, it shows that simple heuristics drawn from geolocated Twitter data offer a good proxy for local commuting patterns; one which outperforms the major existing method for estimating these patterns (the radiation model). Second, it investigates sources of error in the proxy measure, showing that the model performs better on short trips with higher volumes of commuters; it also looks at demographic biases but finds that, surprisingly, measurements are not significantly affected by the fact that the demographic makeup of Twitter users differs significantly from the population as a whole. Finally, it looks at potential ways of going beyond simple heuristics by incorporating temporal information into models.

The number of user reviews of tourist attractions, restaurants, mobile apps, etc. is increasing for all languages; yet, research is lacking on how reviews in multiple languages should be aggregated and displayed. Speakers of different languages may have consistently different experiences, e.g., different information available in different languages at tourist attractions or different user experiences with software due to internationalization/localization choices. This paper assesses the similarity in the ratings given by speakers of different languages to London tourist attractions on TripAdvisor. The correlations between different languages are generally high, but some language pairs are more correlated than others. The results question the common practice of computing average ratings from reviews in many languages.

Multilingualism is common offline, but we have a more limited understanding of the ways multilingualism is displayed online and the roles that multilinguals play in the spread of content between speakers of different languages. We take a computational approach to studying multilingualism using one of the largest user-generated content platforms, Wikipedia. We study multilingualism by collecting and analyzing a large dataset of the content written by multilingual editors of the English, German, and Spanish editions of Wikipedia. This dataset contains over two million paragraphs edited by over 15,000 multilingual users from July 8 to August 9, 2013. We analyze these multilingual editors in terms of their engagement, interests, and language proficiency in their primary and non-primary (secondary) languages and find that the English edition of Wikipedia displays different dynamics from the Spanish and German editions. Users primarily editing the Spanish and German editions make more complex edits than users who edit these editions as a second language. In contrast, users editing the English edition as a second language make edits that are just as complex as the edits by users who primarily edit the English edition. In this way, English serves a special role bringing together content written by multilinguals from many language editions. Nonetheless, language remains a formidable hurdle to the spread of content: we find evidence for a complexity barrier whereby editors are less likely to edit complex content in a second language. In addition, we find that multilinguals are less engaged and show lower levels of language proficiency in their second languages. We also examine the topical interests of multilingual editors and find that there is no significant difference between primary and non-primary editors in each language., Comment: 34 pages, 7 figures

This paper presents a multilingual study on, per single post of microblog text, (a) how much can be said, (b) how much is written in terms of characters and bytes, and (c) how much is said in terms of information content in posts by different organizations in different languages. Focusing on three different languages (English, Chinese, and Japanese), this research analyses Weibo and Twitter accounts of major embassies and news agencies. We first establish our criterion for quantifying "how much can be said" in a digital text based on the openly available Universal Declaration of Human Rights and the translated subtitles from TED talks. These parallel corpora allow us to determine the number of characters and bits needed to represent the same content in different languages and character encodings. We then derive the amount of information that is actually contained in microblog posts authored by selected accounts on Weibo and Twitter. Our results confirm that languages with larger character sets such as Chinese and Japanese contain more information per character than English, but the actual information content contained within a microblog text varies depending on both the type of organization and the language of the post. We conclude with a discussion on the design implications of microblog text limits for different languages., Comment: 9 pages, 4 figures WebSci 2015

This article analyzes users who edit Wikipedia articles about Okinawa, Japan, in English and Japanese. It finds these users are among the most active and dedicated users in their primary languages, where they make many large, high-quality edits. However, when these users edit in their non-primary languages, they tend to make edits of a different type that are overall smaller in size and more often restricted to the narrow set of articles that exist in both languages. Design changes to motivate wider contributions from users in their non-primary languages and to encourage multilingual users to transfer more information across language divides are presented., Comment: In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems, CHI 2015. ACM

Social information is particularly prominent in digital settings where the design of platforms can more easily give real-time information about the behaviour of peers and reference groups and thereby stimulate political activity. Changes to these platforms can generate natural experiments allowing an assessment of the impact of changes in social information and design on participation. This paper investigates the impact of the introduction of trending information on the homepage of the UK government petitions platform. Using interrupted time series and a regression discontinuity design, we find that the introduction of the trending feature had no statistically significant effect on the overall number of signatures per day, but that the distribution of signatures across petitions changes: the most popular petitions gain even more signatures at the expense of those with less signatories. We find significant differences between petitions trending at different ranks, even after controlling for each petition's individual growth prior to trending. The findings suggest a non-negligible group of individuals visit the homepage of the site looking for petitions to sign and therefore see the list of trending petitions, and a significant proportion of this group responds to the social information that it provides. These findings contribute to our understanding of how social information, and the form in which it is presented, affects individual political behaviour in digital settings., Comment: Prepared for delivery at the 2014 Annual Meeting of the American Political Science Association, August 28-31, 2014

This paper maps the national UK web presence on the basis of an analysis of the .uk domain from 1996 to 2010. It reviews previous attempts to use web archives to understand national web domains and describes the dataset. Next, it presents an analysis of the .uk domain, including the overall number of links in the archive and changes in the link density of different second-level domains over time. We then explore changes over time within a particular second-level domain, the academic subdomain .ac.uk, and compare linking practices with variables, including institutional affiliation, league table ranking, and geographic location. We do not detect institutional affiliation affecting linking practices and find only partial evidence of league table ranking affecting network centrality, but find a clear inverse relationship between the density of links and the geographical distance between universities. This echoes prior findings regarding offline academic activity, which allows us to argue that real-world factors like geography continue to shape academic relationships even in the Internet age. We conclude with directions for future uses of web archive resources in this emerging area of research., Comment: To appear in the proceeding of WebSci 2014