Publications (55)
Just Put a Human in the Loop? Investigating LLM-Assisted Annotation for Subjective Tasks
Hope Schroeder, Deb Roy, Jad Kabbara
LLM use in annotation is becoming widespread, and given LLMs' overall promising performance and speed, simply "reviewing" LLM annotations in interpretive tasks can be tempting. In…
M-SENSE: Modeling Narrative Structure in Short Personal Narratives Using Protagonist's Mental Representations
Prashanth Vijayaraghavan, Deb Roy
Narrative is a ubiquitous component of human communication. Understanding its structure plays a critical role in a wide variety of applications, ranging from simple comparative ana…
Bridging Dictionary: AI-Generated Dictionary of Partisan Language Use
Hang Jiang, Doug Beeferman, William Brannon +2
Words often carry different meanings for people from diverse backgrounds. Today's era of social polarization demands that we choose words carefully to prevent miscommunication, esp…
The Structure of Toxic Conversations on Twitter
Martin Saveski, Brandon Roy, Deb Roy
Social media platforms promise to enable rich and vibrant conversations online; however, their potential is often hindered by antisocial behaviors. In this paper, we study the rela…
A Semi-automatic Method for Efficient Detection of Stories on Social Media
Soroush Vosoughi, Deb Roy
Twitter has become one of the main sources of news for many people. As real-world events and emergencies unfold, Twitter is abuzz with hundreds of thousands of stories about the ev…
An AI-Powered Framework for Analyzing Collective Idea Evolution in Deliberative Assemblies
Elinor Poole-Dayan, Deb Roy, Jad Kabbara
In an era of increasing societal fragmentation, political polarization, and erosion of public trust in institutions, representative deliberative assemblies are emerging as a promis…
Automatic Detection and Categorization of Election-Related Tweets
Prashanth Vijayaraghavan, Soroush Vosoughi, Deb Roy
With the rise in popularity of public social media and micro-blogging services, most notably Twitter, the people have found a venue to hear and be heard by their peers without an i…
Conversation Networks
Deb Roy, Lawrence Lessig, Audrey Tang
Picture a community torn over a proposed zoning law. Some are angry, others defensive, and misunderstandings abound. On social media, they broadcast insults at one another; every n…
Perspective-taking to Reduce Affective Polarization on Social Media
Martin Saveski, Nabeel Gillani, Ann Yuan +2
The intensification of affective polarization worldwide has raised new questions about how social media platforms might be further fracturing an already-divided public sphere. As o…
On the Relationship between Truth and Political Bias in Language Models
Suyash Fulay, William Brannon, Shrestha Mohanty +4
Language model alignment research often attempts to ensure that models are not only helpful and harmless, but also truthful and unbiased. However, optimizing these objectives simul…
Divergences in Following Patterns between Influential Twitter Users and Their Audiences across Dimensions of Identity
Suyash Fulay, Nabeel Gillani, Deb Roy
Identity spans multiple dimensions; however, the relative salience of a dimension of identity can vary markedly from person to person. Furthermore, there is often a difference betw…
AI and Collective Decisions: Strengthening Legitimacy and Losers' Consent
Suyash Fulay, Prerna Ravi, Emily Kubin +3
AI is increasingly used to scale collective decision-making, but far less attention has been paid to how such systems can support procedural legitimacy, particularly the conditions…
PersonaLLM: Investigating the Ability of Large Language Models to Express Personality Traits
Hang Jiang, Xiajie Zhang, Xubo Cao +3
Despite the many use cases for large language models (LLMs) in creating personalized chatbots, there has been limited research on evaluating the extent to which the behaviors of pe…
Redrawing attendance boundaries to promote racial and ethnic diversity in elementary schools
Nabeel Gillani, Doug Beeferman, Christine Vega-Pourheydarian +3
Most US school districts draw "attendance boundaries" to define catchment areas that assign students to schools near their homes, often recapitulating neighborhood demographic segr…
RadioTalk: a large-scale corpus of talk radio transcripts
Doug Beeferman, William Brannon, Deb Roy
We introduce RadioTalk, a corpus of speech recognition transcripts sampled from talk radio broadcasts in the United States between October of 2018 and March of 2019. The corpus is…
DeepStance at SemEval-2016 Task 6: Detecting Stance in Tweets Using Character and Word-Level CNNs
Prashanth Vijayaraghavan, Ivan Sysoev, Soroush Vosoughi +1
This paper describes our approach for the Detecting Stance in Tweets task (SemEval-2016 Task 6). We utilized recent advances in short text categorization using deep learning to cre…
Language Models Trained on Media Diets Can Predict Public Opinion
Eric Chu, Jacob Andreas, Stephen Ansolabehere +1
Public opinion reflects and shapes societal behavior, but the traditional survey-based tools to measure it are limited. We introduce a novel approach to probe media diet models --…
Outgroup Animosity Has Risen for Politicians, Journalists, and a Sample of Partisan Users on Twitter and Reddit
Suyash Fulay, Deb Roy
Using language models, we analyze a sample of 67 million tweets and 30 million Reddit comments referencing a set of 215 political entities from 2010-2023 from partisan users, journ…
Audio-Visual Sentiment Analysis for Learning Emotional Arcs in Movies
Eric Chu, Deb Roy
Stories can have tremendous power -- not only useful for entertainment, they can activate our interests and mobilize our actions. The degree to which a story resonates with its aud…
All a-board: sharing educational data science research with school districts
Nabeel Gillani, Doug Beeferman, Cassandra Overney +2
Educational data scientists often conduct research with the hopes of translating findings into lasting change through policy, civil society, or other channels. However, the bridge…
FaciliTrain: Practicing Facilitation Skills through AI-Simulated Group Dialogue
Hang Jiang, Yuanxin Zhu, Diyi Yang +3
Skilled facilitation supports inclusive small-group dialogue, but deliberate practice is hard to scale: it depends on expert coaches, live practice partners, and iterative feedback…
Context is Key: New Approaches to Neural Coherence Modeling
David McClure, Shayne O'Brien, Deb Roy
We formulate coherence modeling as a regression task and propose two novel methods to combine techniques from our setup with pairwise approaches. The first of our methods is a mode…
Computational Analysis of Conversation Dynamics through Participant Responsivity
Margaret Hughes, Brandon Roy, Elinor Poole-Dayan +2
Growing literature explores toxicity and polarization in discourse, with comparatively less work on characterizing what makes dialogue prosocial and constructive. We explore conver…
Interpretable Multi-Modal Hate Speech Detection
Prashanth Vijayaraghavan, Hugo Larochelle, Deb Roy
With growing role of social media in shaping public opinions and beliefs across the world, there has been an increased attention to identify and counter the problem of hate speech…
Bridging Context Gaps: Enhancing Comprehension in Long-Form Social Conversations Through Contextualized Excerpts
Shrestha Mohanty, Sarah Xuan, Jacob Jobraeel +3
We focus on enhancing comprehension in small-group recorded conversations, which serve as a medium to bring people together and provide a space for sharing personal stories and exp…
Prompting Large Language Models with Audio for General-Purpose Speech Summarization
Wonjune Kang, Deb Roy
In this work, we introduce a framework for speech summarization that leverages the processing and reasoning capabilities of large language models (LLMs). We propose an end-to-end s…
Agora: Teaching the Skill of Consensus-Finding with AI Personas Grounded in Human Voice
Prerna Ravi, Om Gokhale, Suyash Fulay +3
Deliberative democratic theory suggests that civic competence: the capacity to navigate disagreement, weigh competing values, and arrive at collective decisions is not innate but d…
Topic Detection and Tracking with Time-Aware Document Embeddings
Hang Jiang, Doug Beeferman, Weiquan Mao +1
The time at which a message is communicated is a vital piece of metadata in many real-world natural language processing tasks such as Topic Detection and Tracking (TDT). TDT system…
Generalized Grounding Graphs: A Probabilistic Framework for Understanding Grounded Commands
Thomas Kollar, Stefanie Tellex, Matthew Walter +8
Many task domains require robots to interpret and act upon natural language commands which are given by people and which refer to the robot's physical surroundings. Such interpreta…
Engaging Politically Diverse Audiences on Social Media
Martin Saveski, Doug Beeferman, David McClure +1
We study how political polarization is reflected in the social media posts used by media outlets to promote their content online. In particular, we track the Twitter posts of sever…
Human Atlas: A Tool for Mapping Social Networks
Martin Saveski, Eric Chu, Soroush Vosoughi +1
Most social network analyses focus on online social networks. While these networks encode important aspects of our lives they fail to capture many real-world connections. Most of t…
Voice to Vision: Enhancing Civic Decision-Making through Co-Designed Data Infrastructure
Maggie Hughes, Cassandra Overney, Ashima Kamra +4
Trust and transparency in civic decision-making processes, like neighborhood planning, are eroding as community members frequently report sending feedback "into a void" without und…
Are Visual Explanations Useful? A Case Study in Model-in-the-Loop Prediction
Eric Chu, Deb Roy, Jacob Andreas
We present a randomized controlled trial for a model-in-the-loop regression task, with the goal of measuring the extent to which (1) good explanations of model predictions increase…
The Empty Chair: Using LLMs to Raise Missing Perspectives in Policy Deliberations
Suyash Fulay, Dimitra Dimitrakopoulou, Deb Roy
Deliberation is essential to well-functioning democracies, yet physical, economic, and social barriers often exclude certain groups, reducing representativeness and contributing to…
Designing Building Blocks for Open-Ended Early Literacy Software
Ivan Sysoev, James H. Gray, Susan Fine +1
English has a convoluted relationship between its pronunciation and spelling, which obscures its phonological structure for early literacy learners. This convoluted relationship ha…
Learning Personas from Dialogue with Attentive Memory Networks
Eric Chu, Prashanth Vijayaraghavan, Deb Roy
The ability to infer persona from dialogue can have applications in areas ranging from computational narrative analysis to personalized dialogue generation. We introduce neural mod…
Generating Black-Box Adversarial Examples for Text Classifiers Using a Deep Reinforced Model
Prashanth Vijayaraghavan, Deb Roy
Recently, generating adversarial examples has become an important means of measuring robustness of a deep learning model. Adversarial examples help us identify the susceptibilities…
ConGraT: Self-Supervised Contrastive Pretraining for Joint Graph and Text Embeddings
William Brannon, Wonjune Kang, Suyash Fulay +4
Learning on text-attributed graphs (TAGs), in which nodes are associated with one or more texts, has been the subject of much recent work. However, most approaches tend to make str…
Annotating the Tweebank Corpus on Named Entity Recognition and Building NLP Models for Social Media Analysis
Hang Jiang, Yining Hua, Doug Beeferman +1
Social media data such as Twitter messages ("tweets") pose a particular challenge to NLP systems because of their short, noisy, and colloquial nature. Tasks such as Named Entity Re…
End-to-End Zero-Shot Voice Conversion with Location-Variable Convolutions
Wonjune Kang, Mark Hasegawa-Johnson, Deb Roy
Zero-shot voice conversion is becoming an increasingly popular research topic, as it promises the ability to transform speech to sound like any speaker. However, relatively little…
Exploring aspects of similarity between spoken personal narratives by disentangling them into narrative clause types
Belen Saldias, Deb Roy
Sharing personal narratives is a fundamental aspect of human social behavior as it helps share our life experiences. We can tell stories and rely on our background to understand th…
LLM Targeted Underperformance Disproportionately Impacts Vulnerable Users
Elinor Poole-Dayan, Deb Roy, Jad Kabbara
While state-of-the-art large language models (LLMs) have shown impressive performance on many tasks, there has been extensive research on undesirable model behavior such as halluci…
Measuring Responsiveness in the Online Public Sphere for the 2016 U.S. Election: Concepts
Pau Perng-Hwa Kung, Deb Roy
The election narrative is formed under the competitions of ideas among critical players involving politicians, news media, public influentials, and the general public. Untangling t…
CommunityLM: Probing Partisan Worldviews from Language Models
Hang Jiang, Doug Beeferman, Brandon Roy +1
As political attitudes have diverged ideologically in the United States, political speech has diverged lingusitically. The ever-widening polarization between the US political parti…
Tweet2Vec: Learning Tweet Embeddings Using Character-level CNN-LSTM Encoder-Decoder
Soroush Vosoughi, Prashanth Vijayaraghavan, Deb Roy
We present Tweet2Vec, a novel method for generating general-purpose vector representation of tweets. The model learns tweet embeddings using character-level CNN-LSTM encoder-decode…
Tweet Acts: A Speech Act Classifier for Twitter
Soroush Vosoughi, Deb Roy
Speech acts are a way to conceptualize speech as action. This holds true for communication on any platform, including social media platforms such as Twitter. In this paper, we expl…
Leveraging Large Language Models for Learning Complex Legal Concepts through Storytelling
Hang Jiang, Xiajie Zhang, Robert Mahari +8
Making legal knowledge accessible to non-experts is crucial for enhancing general legal literacy and encouraging civic participation in democracy. However, legal documents are ofte…
Enhanced Twitter Sentiment Classification Using Contextual Information
Soroush Vosoughi, Helen Zhou, Deb Roy
The rise in popularity and ubiquity of Twitter has made sentiment analysis of tweets an important and well-covered area of research. However, the 140 character limit imposed on twe…
Anonymization of Voices in Spaces for Civic Dialogue: Measuring Impact on Empathy, Trust, and Feeling Heard
Wonjune Kang, Margaret A. Hughes, Deb Roy
Anonymity is a powerful component of many participatory media platforms that can afford people greater freedom of expression and protection from external coercion and interference.…
Applying Large Language Models to Characterize Public Narratives
Elinor Poole-Dayan, Daniel T Kessler, Hannah Chiou +4
Public Narratives (PNs) are key tools for leadership development and civic mobilization, yet their systematic analysis remains challenging due to their subjective interpretation an…
AudienceView: AI-Assisted Interpretation of Audience Feedback in Journalism
William Brannon, Doug Beeferman, Hang Jiang +2
Understanding and making use of audience feedback is important but difficult for journalists, who now face an impractically large volume of audience comments online. We introduce A…
Digital Stylometry: Linking Profiles Across Social Networks
Soroush Vosoughi, Helen Zhou, Deb Roy
There is an ever growing number of users with accounts on multiple social media and networking sites. Consequently, there is increasing interest in matching user accounts and profi…
Me, My Echo Chamber, and I: Introspection on Social Media Polarization
Nabeel Gillani, Ann Yuan, Martin Saveski +2
Homophily -- our tendency to surround ourselves with others who share our perspectives and opinions about the world -- is both a part of human nature and an organizing principle un…
Video SemNet: Memory-Augmented Video Semantic Network
Prashanth Vijayaraghavan, Deb Roy
Stories are a very compelling medium to convey ideas, experiences, social and cultural values. Narrative is a specific manifestation of the story that turns it into knowledge for t…
Expressive Speech Retrieval using Natural Language Descriptions of Speaking Style
Wonjune Kang, Deb Roy
We introduce the task of expressive speech retrieval, where the goal is to retrieve speech utterances spoken in a given style based on a natural language description of that style.…