UBC Computer Science publishes 12 papers in leading Natural Language Processing conference

Several UBC researchers are presenting papers at the Empirical Methods in Natural Language Processing (EMNLP) conference, associated conferences and workshops

If you have ever asked Siri a question, interacted with a chatbot or used a grammar checker, you have reaped the benefits of natural language processing research. As a branch of artificial intelligence, natural language processing involves other diverse fields such as linguistics and human-computer interaction. Researchers studying natural language processing aim to build machines that can generate output resembling human language and how humans communicate in order to create widely applicable tools.

As these language-based tools become more embedded in our daily lives, researchers are studying natural language processing to help these systems become less biased and more robust.

To present their latest research and discuss emerging trends in the field, researchers from the UBC Natural Language Processing Group will be in Suzhou, China from Nov 4-7, 2025, for the Empirical Methods in Natural Language Processing (EMNLP) conference. From a new tool that analyzes bias in text-to-image models to a new method for summarizing multiple documents, UBC researchers have a total of 12 papers accepted at this year’s EMNLP conference.

The following list are the nine accepted papers that will be presented at the main conference:

CEMTM: Contextual Embedding-based Multimodal Topic Modeling
Amirhossein Abaskohi, Raymond Li, Chuyuan Li, Shafiq Joty, Giuseppe Carenini

ChartGaze: Enhancing Chart Understanding in LVLMs with Eye-Tracking Guided Attention Refinement
Ali Salaminian, Amirhossein Abaskohi, Wan-Cyuan Fan, Mir Rayat Imtiaz Hossain, Leonid Sigal, Giuseppe Carenini

Dementia Through Different Eyes: Explainable Modeling of Human and LLM Perceptions for Early Awareness
Lotem Peled-Cohen, Maya Zadok, Nitay Calderon, Hila Gonen, Roi Reichart

FM2DS: Few-Shot Multimodal Multihop Data Synthesis with Knowledge Distillation for Question Answering
Amirhossein Abaskohi, Spandana Gella, Giuseppe Carenini, Issam H. Laradji

LingGym: How Far are LLMs from Thinking Like Field Linguists?
Changbing Yang, Franklin Ma, Freda Shi, Jian Zhu

Mitigate One, Skew Another? Tackling Intersectional Biases in Text-to-Image Models
Pushkar Shukla, Aditya Chinchure, Emily Diana, Alexander Tolbert, Kartik Hosanagar, Vineeth N. Balasubramanian, Leonid Sigal, Matthew A. Turk

OMNIGUARD: An Efficient Approach for AI Safety Moderation Across Modalities
Sahil Verma, Keegan Hines, Jeff Bilmes, Charlotte Siska, Luke Zettlemoyer, Hila Gonen, Chandan Singh

Out of Sight, Not Out of Context? Egocentric Spatial Reasoning in VLMs Across Disjoint Frame
Sahithya Ravi, Gabriel Sarch, Vibhav Vineet, Andrew D Wilson, Balasaravanan Thoravi Kumaravel

Topic-Guided Reinforcement Learning with LLMs for Enhancing Multi-Document Summarization
Chuyuan Li, Austin Xu, Shafiq Joty, Giuseppe Carenini

UBC researchers also have three papers accepted at the associated workshops at EMNLP:

Discourse Relation Recognition with Language Models Under Different Data Availability
Shuhaib Mehri, Chuyuan Li, Giuseppe Carenini
Joint Sixth Workshop on Computational Approaches to Discourse, Context and Document-Level Inferences and Eighth Workshop on Computational Models of Reference, Anaphora and Coreference (CODI-CRAC 2025)

Logically Constrained Decoding
Franklin Ma, Alan J. Hu
The 3rd Workshop on Mathematical Natural Language Processing (Math NLP)

Bridging Information Gaps with Comprehensive Answers: Improving the Diversity and Informativeness of Follow-Up Questions
Zhe Liu, Taekyu Kang, Haoyu Wang, Seyed Hossein Alavim, Vered Shwartz
The 14th Joint Conference on Lexical and Computational Semantics (SEM 2025)