PairBench: A Systematic Framework for Selecting Reliable Judge VLMs Permalink
Aarash Feizi, Sai Rajeswar, Adriana Romero-Soriano, Reihaneh Rabbany, Spandana Gella, Valentina Zantedeschi, Joao Monteiro
Posts by Category
- Publications 120
- Teaching 11
- TGA 5
- information-integrity 5
- TGB 5
- online-crime 4
- online-toxicity 3
- ideology-and-polarization 3
- social-simulations 1
Publications
AlignVLM: Bridging Vision and Language Latent Spaces for Multimodal Understanding Permalink
Ahmed Masry, Juan A. Rodriguez, Tianyu Zhang, Suyuchen Wang, Chao Wang, Aarash Feizi, Akshay Kalkunte Suresh, Abhay Puri, Xiangru Jian, Pierre-Andr’e Noel, S...
Online Influence Campaigns: Strategies and Vulnerabilities Permalink
Andreea Musulan, Veronica Xia, Ethan Kosak-Hine, Tom Gibbs, Vidya Sujaya, Reihaneh Rabbany, J. Godbout, Kellin Pelrine
BigDocs: An Open and Permissively-Licensed Dataset for Training Multimodal Models on Document and Code Tasks Permalink
Juan Rodriguez, Xiangru Jian, Siba Smarak Panigrahi, Tianyu Zhang, Aarash Feizi, Abhay Puri, Akshay Kalkunte, Franccois Savard, Ahmed Masry, Shravan Nayak, R...
Epistemic Integrity in Large Language Models Permalink
Bijean Ghafouri, Shahrad Mohammadzadeh, James Zhou, Pratheeksha Nair, Jacob-Junqi Tian, Mayank Goel, Reihaneh Rabbany, J. Godbout, Kellin Pelrine
A Guide to Misinformation Detection Datasets Permalink
Data is a barrier to reliable misinformation detection solutions. To address this, we curated the largest collection of (mis)information datasets in the lite...
FairLoRA: Unpacking Bias Mitigation in Vision Models with Fairness-Driven Low-Rank Adaptation Permalink
Rohan Sukumaran, Aarash Feizi, Adriana Romero-Sorian, G. Farnadi Paper Abstrac...
Hallucination Detox: Sensitive Neuron Dropout (SeND) for Large Language Model Training Permalink
Shahrad Mohammadzadeh, Juan David Guerra, Marco Bonizzato, Reihaneh Rabbany, G. Farnadi Paper ...
A Simulation System Towards Solving Societal-Scale Manipulation Permalink
In this paper, we present a multi-agent simulator based on Deepmind’s Concordia, a software library for LLM-based multi-agent simulations of real world human...
Emerging Vulnerabilities in Frontier Models: Multi-Turn Jailbreak Attacks Permalink
Tom Gibbs, Ethan Kosak-Hine, George Ingebretsen, Jason Zhang, Julius Broomfield, Sara Pieri, Reihaneh Iranmanesh, Reihaneh Rabbany, Kellin Pelrine
Web Retrieval Agents for Evidence-Based Misinformation Detection Permalink
We demonstrated an effective two LLM agent architecture for misinformation detection and fact-checking. It can increase the macro F1 of misinformation detect...
Data Poisoning in LLMs: Jailbreak-Tuning and Scaling Laws Permalink
Dillon Bowen, Brendan Murphy, Will Cai, David Khachaturov, A. Gleave, Kellin Pelrine Paper ...
UTG: Towards a Unified View of Snapshot and Event Based Models for Temporal Graphs Permalink
Here we introduce the Unified Temporal Graph (UTG) framework, which integrates snapshot-based and event-based machine learning models for temporal graphs und...
Regional and Temporal Patterns of Partisan Polarization during the COVID-19 Pandemic in the United States and Canada Permalink
The COVID-19 pandemic sparked rigorous debate about public health measures online, as the timing and extent of interventions unfolded. This project evaluates...
Game On, Hate Off: A Study of Toxicity in Online Multiplayer Environments Permalink
While game companies are addressing the call to reduce toxicity and promote player health, the need to understand toxicity trends across time is important. W...
Can Go AIs be adversarially robust? Permalink
Tom Tseng, Euan McLean, Kellin Pelrine, T. T. Wang, A. Gleave
TGB 2.0: A Benchmark for Learning on Temporal Knowledge Graphs and Heterogeneous Graphs Permalink
TGB 2.0 introduces a novel benchmarking framework for evaluating methods for predicting future links on Temporal Knowledge Graphs (TKGs) and Temporal Heterog...
Static graph approximations of dynamic contact networks for epidemic forecasting Permalink
In this work, we introduce new, more rigorous evaluation procedures for link prediction in dynamic graphs, addressing the challenges and real-world considera...
MiniMol: A Parameter-Efficient Foundation Model for Molecular Learning Permalink
Kerstin Kläser, Bla.zej Banaszewski, S. Maddrell-Mander, Callum McLean, Luis Müller, Alipanah Parviz, Shenyang Huang, Andrew W. Fitzgibbon
T-NET: Weakly Supervised Graph Learning for Combatting Human Trafficking Permalink
Detecting suspicious ads is challenging due to the sensitive, complex, and unlabeled nature of the data. T-Net addresses this as a weakly supervised graph le...
Temporal Graph Analysis with TGX Permalink
Here we introduce TGX, a Python package specifically designed for the analysis of temporal networks, addressing a gap in existing software libraries that pri...
Scavenging Hyena: Distilling Transformers into Long Convolution Models Permalink
Tokiniaina Raharison Ralambomihanta, Shahrad Mohammadzadeh, Mohammad Sami Nur Islam, Wassim Jabbour, Laurence Liang
Combining Confidence Elicitation and Sample-based Methods for Uncertainty Quantification in Misinformation Mitigation Permalink
LLMs struggle with hallucinations and overconfident predictions. Uncertainty quantification can improve their reliability and helpfulness. We proposed an unc...
Comparing GPT-4 and Open-Source Language Models in Misinformation Mitigation Permalink
Tyler Vergho, J. Godbout, Reihaneh Rabbany, Kellin Pelrine
GPS-SSL: Guided Positive Sampling to Inject Prior Into Self-Supervised Learning Permalink
Aarash Feizi, Randall Balestriero, Adriana Romero-Soriano, Reihaneh Rabbany
Uncertainty Resolution in Misinformation Detection Permalink
Yury Orlovskiy, Camille Thibault, Anne Imouza, J. Godbout, Reihaneh Rabbany, Kellin Pelrine
Quantifying learning-style adaptation in effectiveness of LLM teaching Permalink
Ruben Weijers, Gabrielle Fidelis de Castilho, J. Godbout, Reihaneh Rabbany, Kellin Pelrine
An Evaluation of Language Models for Hyperpartisan Ideology Detection in Persian Twitter Permalink
While LLMs have demonstrated how effective they can be for tasks in the English language, such as detecting social media users’ political ideology, their per...
Exploiting Novel GPT-4 APIs Permalink
Kellin Pelrine, Mohammad Taufeeque, Michal Zajkac, Euan McLean, A. Gleave
Exhaustive Evaluation of Dynamic Link Prediction Permalink
Farimah Poursafaei, Reihaneh Rabbany
SWEET : Weakly Supervised Person Name Extraction for Fighting Human Trafficking Permalink
SWEET is a weak supervision pipeline for extracting person names from noisy escort ads, addressing the challenge of limited labeled data. SWEET combines rule...
Towards Detecting Contextual Real-Time Toxicity for In-Game Chat Permalink
A simple and scalable model that reliably detects toxic content in real-time for a line of chat by including chat history and metadata. ToxBuster consistentl...
Towards Foundational Models for Molecular Learning on Large-Scale Multi-Task Datasets Permalink
D. Beaini, Shenyang Huang, Joao Alex Cunha, Gabriela Moisescu-Pareja, Oleksandr Dymov, Sam Maddrell-Mander, Callum McLean, Frederik Wenkel, Luis Muller, Jama...
Party Prediction for Twitter Permalink
Partisanship has increasingly become a major point of contention in public discourse online, and as a result, researchers have developed a variety of methods...
Open, Closed, or Small Language Models for Text Classification? Permalink
Hao Yu, Zachary Yang, Kellin Pelrine, J. Godbout, Reihaneh Rabbany
Towards Temporal Edge Regression: A Case Study on Agriculture Trade Between Nations Permalink
Lekang Jiang, Caiqi Zhang, Farimah Poursafaei, Shenyang Huang
Temporal Graph Benchmark for Machine Learning on Temporal Graphs Permalink
In this paper, we introduce the Temporal Graph Benchmark (TGB), a comprehensive collection of large-scale, diverse datasets designed for the realistic, repro...
Graph learning with programmatic weak supervision Permalink
Pratheeksha Nair
Better Bridges Between Model and Real World Permalink
Kellin Pelrine
Towards Reliable Misinformation Mitigation: Generalization, Uncertainty, and GPT-4 Permalink
We propose focusing on generalization, uncertainty, and how to leverage recent large language models, in order to create more practical tools to evaluate inf...
ToxBuster: In-game Chat Toxicity Buster with BERT Permalink
Zachary Yang, Yasmine Maricar, M. Davari, Nicolas Grenon-Godbout, Reihaneh Rabbany
Fast and Attributed Change Detection on Dynamic Graphs with Density of States Permalink
We introduce a novel spectral method called Scalable Change Point Detection (SCPD) to address the limitations of current solutions in detecting anomalous cha...
Social Media as a Vector for Escort Ads:A Study on OnlyFans advertisements on Twitter Permalink
Maricarmen Arenas, Pratheeksha Nair, Reihaneh Rabbany, G. Farnadi
DisKeyword: Tweet Corpora Exploration for Keyword Selection Permalink
Sacha Lévy, Reihaneh Rabbany
DeltaShield: Information Theory for Human- Trafficking Detection Permalink
Catalina Vajiac, Meng-Chieh Lee, Aayushi Kulshrestha, Sacha Lévy, Namyong Park, Andreas M. Olligschlaeger, Cara Jones, Reihaneh Rabbany, C. Faloutsos
GPS++: Reviving the Art of Message Passing for Molecular Property Prediction Permalink
Dominic Masters, Josef Dean, K. Klaser, Zhiyi Li, Sam Maddrell-Mander, Adam Sanders, Hatem Helal, D. Beker, A. Fitzgibbon, Shenyang Huang, Ladislav Rampášek,...
Laplacian Change Point Detection for Single and Multi-view Dynamic Graphs Permalink
Shenyang Huang, Samy Coulombe, Yasmeen Hitti, Reihaneh Rabbany, Guillaume Rabusseau
SWEET - Weakly Supervised Person Name Extraction for Fighting Human Trafficking Permalink
Javin Liu, Hao Yu, Vidya Sujaya, Pratheeksha Nair, Kellin Pelrine, Reihaneh Rabbany
Unveiling Identity Biases in Toxicity Detection : A Game-Focused Dataset and Reactivity Analysis Approach Permalink
Identity biases arise commonly from annotated datasets, can be propagated in language models and can cause further harm to marginal groups. Existing bias ben...
Adversarial Policies Beat Superhuman Go AIs Permalink
T. T. Wang, A. Gleave, Nora Belrose, Tom Tseng, Joseph Miller, Kellin Pelrine, Michael Dennis, Yawen Duan, V. Pogrebniak, S. Levine, Stuart Russell
TRAFFIC BOARD : Digital Spatio-Temporal Pinboard for Human Trafficking Detection Permalink
Catalina Vajiac, Duen Horng, †. Chau, Andreas M. Olligschlaeger, Pratheeksha Nair, Meng-Chieh Lee, M. Cazzolato, Reihaneh Rabbany, Christos Faloutsos
TrafficVis: Visualizing Organized Activity and Spatio-Temporal Patterns for Detecting and Labeling Human Trafficking Permalink
TrafficVis is an interactive interface for detecting and labeling human trafficking (HT) in clusters of escort ads. Developed with domain experts, it uses ad...
Active Keyword Selection to Track Evolving Topics on Twitter Permalink
Sacha L’evy, Farimah Poursafaei, Kellin Pelrine, Reihaneh Rabbany
Revisiting Hotels-50K and Hotel-ID Permalink
Aarash Feizi, Arantxa Casanova, Adriana Romero-Soriano, Reihaneh Rabbany
Towards Better Evaluation for Dynamic Link Prediction Permalink
In this work, we introduce new, more rigorous evaluation procedures for link prediction in dynamic graphs, addressing the challenges and real-world considera...
PCR-based Evaluation of Human Papillomavirus Genotypes in Oral Lichen Planus Permalink
A. Vijayan, Arvind Muthukrishnan, Aparna Nair, S. Fathima, Pratheeksha Nair, J. Roshan
VisPaD: Visualization and Pattern Discovery for Fighting Human Trafficking Permalink
Pratheeksha Nair, Yifei Li, Catalina Vajiac, Andreas M. Olligschlaeger, Meng-Chieh Lee, Namyong Park, Duen Horng Chau, C. Faloutsos, Reihaneh Rabbany, Chieh ...
The Winners and Losers of Rental Tribunals Permalink
Andreea Musulan
Extracting Person Names from User Generated Text: Named-Entity Recognition for Combating Human Trafficking Permalink
Yifei Li, Pratheeksha Nair, Kellin Pelrine, Reihaneh Rabbany
A Strong Node Classification Baseline for Temporal Graphs Permalink
Here we introduce a robust and effective baseline method for node classification in temporal graphs, serving as a benchmark for evaluating more complex model...
Curating the Twitter Election Integrity Datasets for Better Online Troll Characterization Permalink
Albert Orozco, Reihaneh Rabbany
Online Partisan Polarization of COVID-19 Permalink
Zachary Yang, Anne Imouza, Kellin Pelrine, Sacha Lévy, Jiewen Liu, Gabrielle Desrosiers-Brisebois, J. Godbout, A. Blais, Reihaneh Rabbany
Understanding Capacity Saturation in Incremental Learning Permalink
Shenyang Huang, Vincent François-Lavet, Guillaume Rabusseau
Graph Attention Networks with Positional Embeddings Permalink
Liheng Ma, Reihaneh Rabbany, Adriana Romero-Soriano
The Surprising Performance of Simple Baselines for Misinformation Detection Permalink
While many sophisticated detection models have been proposed in the literature, they were often compared with older NLP baselines such as SVMs, CNNs, and LST...
INFOSHIELD: Generalizable Information-Theoretic Human-Trafficking Detection Permalink
INFOSHIELD is a scalable, parameter-free, and interpretable tool for detecting near-duplicate document clusters, with applications in human trafficking detec...
Efficient detection of online communities and social bot activity during electoral campaigns Permalink
L. Rheault, Andreea Musulan
RAFFIC V IS : Fighting Human Trafficking through Visualization Permalink
Catalina Vajiac, Andreas M. Olligschlaeger, Yifei Li, Pratheeksha Nair, Meng-Chieh Lee, Namyong Park, Reihaneh Rabbany, Duen Horng Chau, C. Faloutsos
Few Shot Image Generation via Implicit Autoencoding of Support Sets Permalink
Shenyang Huang, Kuan-Chieh Jackson Wang, Alireza Makhzani
Scalable Change Point Detection for Dynamic Graphs Permalink
Shenyang Huang, Guillaume Rabusseau, Reihaneh Rabbany
SigTran: Signature Vectors for Detecting Illicit Activities in Blockchain Transaction Networks Permalink
SigTran is an efficient graph-based method for identifying illicit nodes in blockchain networks. SigTran constructs a graph from blockchain transaction recor...
Graph Neural Networks Learn Twitter Bot Behaviour Permalink
Albert Orozco, Sacha Lévy, Reihaneh Rabbany
ComplexDataLab at W-NUT 2020 Task 2: Detecting Informative COVID-19 Tweets by Attending over Linked Documents Permalink
Kellin Pelrine, Jacob Danovitch, Albert Orozco Camacho, Reihaneh Rabbany
Contact Graph Epidemic Modelling of COVID-19 for Transmission and Intervention Strategies Permalink
Abby Leung, Xiaoye Ding, Shenyang Huang, Reihaneh Rabbany
Incorporating dynamic flight network in SEIR to model mobility between populations Permalink
In this paper, we introduce a significant enhancement to the standard SEIR (Susceptible, Exposed, Infectious, Recovered) epidemiological model by integrating...
Structure Aware Negative Sampling in Knowledge Graphs Permalink
Kian Ahrabian, Aarash Feizi, Yasmin Salehi, William L. Hamilton, A. Bose
Global Surveillance of COVID-19 by mining news media using a multi-source dynamic embedded topic model Permalink
Yue Li, Pratheeksha Nair, Zhi Wen, I. Chafi, A. Okhmatovskaia, G. Powell, Yannan Shen, D. Buckeridge
Explaining Support for COVID-19 Cell Phone Contact Tracing Permalink
L. Rheault, Andreea Musulan
Laplacian Change Point Detection for Dynamic Graphs Permalink
Shenyang Huang, Yasmeen Hitti, Guillaume Rabusseau, Reihaneh Rabbany Knowledge Discovery and Data Mining Paper ...
Group Equivariant Deep Reinforcement Learning Permalink
Arnab Kumar Mondal, Pratheeksha Nair, Kaleem Siddiqi
RandomNet: Towards Fully Automatic Neural Architecture Design for Multimodal Learning Permalink
S. Alletto, Shenyang Huang, Vincent François-Lavet, Yohei Nakata, Guillaume Rabusseau
Investigating the Role of Social Bots During the 2019 Canadian Election Permalink
L. Rheault, Andreea Musulan
Machine learning analysis of exome trios to contrast the genomic architecture of autism and schizophrenia Permalink
Sameer Sardaar, B. Qi, Alexandre Dionne‐Laporte, G. Rouleau, Reihaneh Rabbany, Y. Trakadis
Scaling up Simhash Permalink
Rameshwar Pratap, Anup Anand Deshmukh, Pratheeksha Nair, Anirudh Ravi
A Study on Active Learning for Graphs Permalink
Pratheeksha Nair, Zhi Wen
SCG: Spotting Coordinated Groups in Social Media Permalink
Junhao Wang, Sacha Lévy, Ren Wang, Aayushi Kulshrestha, Guillaume Rabusseau, Reihaneh Rabbany
SGP: Spotting Groups Polluting the Online Political Discourse Permalink
Junhao Wang, Sacha Lévy, Ren Wang, Aayushi Kulshrestha, Reihaneh Rabbany
Neural Architecture Search for Class-incremental Learning Permalink
Shenyang Huang, Vincent François-Lavet, Guillaume Rabusseau
Anomaly Detection with Joint Representation Learning of Content and Connection Permalink
Junhao Wang, Renhao Wang, Aayushi Kulshrestha, Reihaneh Rabbany
Politicians in the line of fire: Incivility and the treatment of women on social media Permalink
L. Rheault, Erica Rayment, Andreea Musulan
A Faster Sampling Algorithm for Spherical k-means Permalink
Rameshwar Pratap, Anup Anand Deshmukh, Pratheeksha Nair, T. Dutt
A Scalable Clustering Algorithm for Serendipity in Recommender Systems Permalink
Anup Anand Deshmukh, Pratheeksha Nair, Shrisha Rao
Social-Affiliation Networks: Patterns and the SOAR Model Permalink
D. Eswaran, Reihaneh Rabbany, A. Dubrawski, C. Faloutsos
Active Search of Connections for Case Building and Combating Human Trafficking Permalink
Reihaneh Rabbany, David Bayani, A. Dubrawski
Modular Networks for Validating Community Detection Algorithms Permalink
Justin Fagnan, Afra Abnar, Reihaneh Rabbany, Osmar R Zaiane
A Note on the Unconditional Bias of the Nadaraya-Watson Regression Estimator Permalink
Kellin Pelrine
Assortativity: Proclivity Index for Attributed Networks (P RO N E ) Permalink
Reihaneh Rabbany, D. Eswaran, Artur W. Dubrawski, Christos Faloutsos
Beyond Assortativity: Proclivity Index for Attributed Networks (ProNe) Permalink
Reihaneh Rabbany, D. Eswaran, A. Dubrawski, C. Faloutsos
A General Clustering Agreement Index: For Comparing Disjoint and Overlapping Clusters Permalink
Reihaneh Rabbany, Osmar R Zaiane
PROCLIVITY PATTERNS IN ATTRIBUTED GRAPHS Permalink
Reihaneh Rabbany, D. Eswaran, C. Faloutsos, A. Dubrawski
Generalization of clustering agreements and distances for overlapping clusters and network communities Permalink
Reihaneh Rabbany, Osmar R Zaiane
Evaluation of Community Mining Algorithms in the Presence of Attributes Permalink
Reihaneh Rabbany, Osmar R Zaiane
Generating Attributed Networks with Communities Permalink
C. Largeron, Pierre-Nicolas Mougel, Reihaneh Rabbany, Osmar R Zaiane
Community Dynamics: Event and Role Analysis in Social Network Analysis Permalink
Justin Fagnan, Reihaneh Rabbany, M. Takaffoli, Eric Verbeek, Osmar R Zaiane
Generalization of clustering agreements and distances for overlapping clusters and network communities Permalink
Reihaneh Rabbany, Osmar R Zaiane
Community evolution prediction in dynamic social networks Permalink
M. Takaffoli, Reihaneh Rabbany, Osmar R Zaiane
SSRM: structural social role mining for dynamic social networks Permalink
Afra Abnar, M. Takaffoli, Reihaneh Rabbany, Osmar R Zaiane
Augmentative Message Passing for Traveling Salesman Problem and Graph Partitioning Permalink
Siamak Ravanbakhsh, Reihaneh Rabbany, R. Greiner
Mining Large Scale Data from National Educational Achievement Tests A Case Study Permalink
Reihaneh Rabbany, Osmar R Zaiane, Samira ElAtia
Collaborative Learning of Students in Online Discussion Forums: A Social Network Analysis Perspective Permalink
Reihaneh Rabbany, Samira ElAtia, M. Takaffoli, Osmar R Zaiane
Communities validity: methodical evaluation of community mining algorithms Permalink
Reihaneh Rabbany, M. Takaffoli, Justin Fagnan, Osmar R Zaiane, R. Campello
Incremental local community identification in dynamic social networks Permalink
M. Takaffoli, Reihaneh Rabbany, Osmar R Zaiane
Relative Validity Criteria for Community Mining Algorithms Synonyms Evaluation Permalink
Reihaneh Rabbany, Mansoreh Takaffoli, Justin Fagnan, Osmar R. Zäıane, R. Campello
Relative Validity Criteria for Community Mining Algorithms Permalink
Reihaneh Rabbany, M. Takaffoli, Justin Fagnan, Osmar R Zaiane, R. Campello
Social network analysis and mining to support the assessment of on-line student participation Permalink
Reihaneh Rabbany, M. Takaffoli, Osmar R Zaiane
A Diffusion of Innovation-Based Closeness Measure for Network Associations Permalink
Reihaneh Rabbany, Osmar R Zaiane
Web service matching for RESTful web services Permalink
Reihaneh Rabbany, Eleni Stroulia, Osmar R Zaiane
Analyzing Participation of Students in Online Courses Using Social Network Analysis Techniques Permalink
Reihaneh Rabbany, M. Takaffoli, Osmar R Zaiane
Meerkat: Community Mining with Dynamic Social Networks Permalink
Jiyang Chen, Justin Fagnan, R. Goebel, Reihaneh Rabbany, Farzad Sangi, M. Takaffoli, Eric Verbeek, Osmar R Zaiane
Teaching
Applied Machine Learning
Course codes: COMP 551 (Winter 2025) Instructors: Reihaneh Rabbany Location: Stewart Biology Building S1/4 (lectures will be recorded) Time: Tuesday...
Network Science
Course codes: COMP 599 (Fall 2024) Instructors: Reihaneh Rabbany
Applied Machine Learning
Course codes: COMP 551 (Winter 2025) Instructors: Reihaneh Rabbany, Isabeau Prémont-Schwarz Location: McConnell 204 Time: Mondays and Wednesdays, 14...
Applied Machine Learning
Course codes: COMP 551 (Winter 2023) Instructors: Reihaneh Rabbany Location: Stewart Biology Building S1/4 (lectures will be recorded) Time: Tuesday...
Network Science
Course codes: COMP 599 (Fall 2022) Instructors: Reihaneh Rabbany Classroom: Rutherford Physics Building 114 Time: Tuesdays and Thursdays, 10:05 am -...
Applied Machine Learning
Course codes: COMP 551 (Winter 2022) Instructors: Reihaneh Rabbany Location: Remote (lectures will be recorded) Time: Tuesdays and Thursdays, 1:00 p...
Network Science
Course codes: COMP 599 (Fall 2021) Instructors: Reihaneh Rabbany Classroom: Macdonald Engineering Building 276 [remote participation through zoom in M...
Applied Machine Learning
Course codes: COMP 551 (Winter 2021) Instructors: Reihaneh Rabbany Location: Remote through MyCourse Course Website: here
Network Science
Course codes: COMP 596 (Fall 2020) Instructors: Reihaneh Rabbany Location: Remote through MyCourses Time: Monday & Wednesday, 10:00 am – 11:30 p...
Applied Machine Learning
Course codes: COMP 551 (Winter 2020) Instructors: Reihaneh Rabbany Classroom: Strathcona Anatomy & Dentistry M-1 Time: Tuesdays and Thursdays, 1...
Network Science
Course codes: COMP 596 (Fall 2019) Instructors: Reihaneh Rabbany Classroom: ENGMC 103 Time: Tuesdays and Thursdays, 10:00 am – 11:30 pm Course Web...
TGA
Static graph approximations of dynamic contact networks for epidemic forecasting Permalink
In this work, we introduce new, more rigorous evaluation procedures for link prediction in dynamic graphs, addressing the challenges and real-world considera...
Fast and Attributed Change Detection on Dynamic Graphs with Density of States Permalink
We introduce a novel spectral method called Scalable Change Point Detection (SCPD) to address the limitations of current solutions in detecting anomalous cha...
A Strong Node Classification Baseline for Temporal Graphs Permalink
Here we introduce a robust and effective baseline method for node classification in temporal graphs, serving as a benchmark for evaluating more complex model...
SigTran: Signature Vectors for Detecting Illicit Activities in Blockchain Transaction Networks Permalink
SigTran is an efficient graph-based method for identifying illicit nodes in blockchain networks. SigTran constructs a graph from blockchain transaction recor...
Incorporating dynamic flight network in SEIR to model mobility between populations Permalink
In this paper, we introduce a significant enhancement to the standard SEIR (Susceptible, Exposed, Infectious, Recovered) epidemiological model by integrating...
information-integrity
A Guide to Misinformation Detection Datasets Permalink
Data is a barrier to reliable misinformation detection solutions. To address this, we curated the largest collection of (mis)information datasets in the lite...
Web Retrieval Agents for Evidence-Based Misinformation Detection Permalink
We demonstrated an effective two LLM agent architecture for misinformation detection and fact-checking. It can increase the macro F1 of misinformation detect...
Combining Confidence Elicitation and Sample-based Methods for Uncertainty Quantification in Misinformation Mitigation Permalink
LLMs struggle with hallucinations and overconfident predictions. Uncertainty quantification can improve their reliability and helpfulness. We proposed an unc...
Towards Reliable Misinformation Mitigation: Generalization, Uncertainty, and GPT-4 Permalink
We propose focusing on generalization, uncertainty, and how to leverage recent large language models, in order to create more practical tools to evaluate inf...
The Surprising Performance of Simple Baselines for Misinformation Detection Permalink
While many sophisticated detection models have been proposed in the literature, they were often compared with older NLP baselines such as SVMs, CNNs, and LST...
TGB
UTG: Towards a Unified View of Snapshot and Event Based Models for Temporal Graphs Permalink
Here we introduce the Unified Temporal Graph (UTG) framework, which integrates snapshot-based and event-based machine learning models for temporal graphs und...
TGB 2.0: A Benchmark for Learning on Temporal Knowledge Graphs and Heterogeneous Graphs Permalink
TGB 2.0 introduces a novel benchmarking framework for evaluating methods for predicting future links on Temporal Knowledge Graphs (TKGs) and Temporal Heterog...
Temporal Graph Analysis with TGX Permalink
Here we introduce TGX, a Python package specifically designed for the analysis of temporal networks, addressing a gap in existing software libraries that pri...
Temporal Graph Benchmark for Machine Learning on Temporal Graphs Permalink
In this paper, we introduce the Temporal Graph Benchmark (TGB), a comprehensive collection of large-scale, diverse datasets designed for the realistic, repro...
Towards Better Evaluation for Dynamic Link Prediction Permalink
In this work, we introduce new, more rigorous evaluation procedures for link prediction in dynamic graphs, addressing the challenges and real-world considera...
online-crime
T-NET: Weakly Supervised Graph Learning for Combatting Human Trafficking Permalink
Detecting suspicious ads is challenging due to the sensitive, complex, and unlabeled nature of the data. T-Net addresses this as a weakly supervised graph le...
SWEET : Weakly Supervised Person Name Extraction for Fighting Human Trafficking Permalink
SWEET is a weak supervision pipeline for extracting person names from noisy escort ads, addressing the challenge of limited labeled data. SWEET combines rule...
TrafficVis: Visualizing Organized Activity and Spatio-Temporal Patterns for Detecting and Labeling Human Trafficking Permalink
TrafficVis is an interactive interface for detecting and labeling human trafficking (HT) in clusters of escort ads. Developed with domain experts, it uses ad...
INFOSHIELD: Generalizable Information-Theoretic Human-Trafficking Detection Permalink
INFOSHIELD is a scalable, parameter-free, and interpretable tool for detecting near-duplicate document clusters, with applications in human trafficking detec...
online-toxicity
Game On, Hate Off: A Study of Toxicity in Online Multiplayer Environments Permalink
While game companies are addressing the call to reduce toxicity and promote player health, the need to understand toxicity trends across time is important. W...
Towards Detecting Contextual Real-Time Toxicity for In-Game Chat Permalink
A simple and scalable model that reliably detects toxic content in real-time for a line of chat by including chat history and metadata. ToxBuster consistentl...
Unveiling Identity Biases in Toxicity Detection : A Game-Focused Dataset and Reactivity Analysis Approach Permalink
Identity biases arise commonly from annotated datasets, can be propagated in language models and can cause further harm to marginal groups. Existing bias ben...
ideology-and-polarization
Regional and Temporal Patterns of Partisan Polarization during the COVID-19 Pandemic in the United States and Canada Permalink
The COVID-19 pandemic sparked rigorous debate about public health measures online, as the timing and extent of interventions unfolded. This project evaluates...
An Evaluation of Language Models for Hyperpartisan Ideology Detection in Persian Twitter Permalink
While LLMs have demonstrated how effective they can be for tasks in the English language, such as detecting social media users’ political ideology, their per...
Party Prediction for Twitter Permalink
Partisanship has increasingly become a major point of contention in public discourse online, and as a result, researchers have developed a variety of methods...
social-simulations
A Simulation System Towards Solving Societal-Scale Manipulation Permalink
In this paper, we present a multi-agent simulator based on Deepmind’s Concordia, a software library for LLM-based multi-agent simulations of real world human...