Attending Midwest Speech and Language Days (MSLD) 2026 hosted by University of Illinois Urbana-Champaign NLP. This is my 3rd time attending at MSLD and representing @purduecs.bsky.social Excited to present our work.
#msld2026 #msld #nlproc
Preprint: arxiv.org/pdf/2601.17172
Extended abstract titled “Who Gets Which Message? Auditing Demographic Bias in LLM-Generated Targeted Text” has been accepted to @ic2s2.bsky.social
Excited to present my work & meet awesome researchers @vcsi.bsky.social
@purduecs.bsky.social @uvmvermont.bsky.social
#nlproc #css #ic2s2
Pre-print 2: arxiv.org/pdf/2604.07562
We propose a post-hoc framework that uses LLMs as semantic judges to refine unsupervised text clusters via coherence verification, redundancy resolution, and label grounding—improving both coherence and human alignment without supervision.
Pre-print 1: arxiv.org/pdf/2601.17172
Who gets the “stronger” message?
In this paper, we show that LLMs don’t just personalize—they shift tone, style, and persuasion based on demographics, often amplifying stereotypes in targeted communication.
Excited to share that two of my long papers are accepted to @aclmeeting.bsky.social ! 🎉
1. Who Gets Which Message? Auditing Demographic Bias in LLM-Generated Targeted Text
2. Reasoning-Based Refinement of Unsupervised Text Clusters with LLMs
#ACL2026 #NLProc #LLMs #ACL26 @purduecs.bsky.social
Huge thanks to the IEEE DISTILL 2025 organizers for inviting me to give a talk! Loved sharing ideas and connecting with brilliant minds shaping the future of trustworthy AI #DISTILL2025 #AI #LLMs #ResponsibleAI #IEEE
Paper link: aclanthology.org/2025.finding...
Presenting our work titled “Post-hoc Study of Climate Microtargeting on Social Media Ads with LLMs: Thematic Insights and Fairness Evaluation” @emnlpmeeting.bsky.social today. Please feel free to join. #emnlp #EMNLP2025 #nlproc
Glad to serve as a tutorial co-chair of @icwsm.bsky.social again!
Please submit your tutorial proposal by January 15, 2026. #ICWSM2026 #ICWSM
CfP: www.icwsm.org/2026/submit....
In summary, our method provides an explainable way to reverse-engineer targeting practices and uncover potential bias or messaging disparities. Looking forward to connecting with the community around Ethics, Bias & Fairness in NLP #Ethics #Bias #Fairness #FairAI #AIethics #EMNLP
💡 Findings:
• Climate ads use distinct themes: activism for young adults, caregiving & advocacy for women, economic/innovation frames for men
• While LLMs perform well, biases persist, especially for males
• Insights: need for more inclusive, transparent, & accountable climate communication
We leveraged LLMs to audit climate ad microtargeting on social media, proposing a third-party, post-hoc framework to:
1️⃣ Identify which demographics (age, gender) ads are targeting
2️⃣ Generate transparent explanations that reveal messaging strategies
3️⃣ Evaluate fairness using established metrics
This work builds on one of the future directions I proposed during my Ph.D. defense, making the acceptance especially meaningful. 🌟
Paper link: arxiv.org/pdf/2410.05401
🎉Happy to share that our long paper titled "Post-hoc Study of Climate Microtargeting on Social Media Ads with LLMs: Thematic Insights and Fairness Evaluation" has been accepted to #EMNLP2025 findings. Hoping to work more in Ethics, Bias, and Fairness track #emnlp25 #nlproc
@emnlpmeeting.bsky.social
Slide: tunazislam.github.io/files/Latent...
Paper link: ojs.aaai.org/index.php/IC...
#machineintheloop #llmsintheloop #latentthemesdiscovery #climatecampaigns #vaccinecampaigns #socialmedia #computationalsocialscience
First page of the slide
Presentation was going well after technical glitch
My ICWSM 2025 badge
Beautiful campus- Copenhagen University
Kicking off 1st day @icwsm.bsky.social at beautiful Copenhagen University campus 🇩🇰. Done with presentation of our long paper titled “Discovering Latent Themes in Social Media Messaging: A Machine-in-the-Loop Approach Integrating LLMs” #icwsm #icwsm2025 #icwsm25 #nlproc #PurdueCS #PurdueUniversity
Paper title “Discovering Latent Themes in Social Media Messaging: A Machine-in-the-Loop Approach Integrating LLMs
Authors”
Paper link: ojs.aaai.org/index.php/IC...
Staying Copenhagen, Denmark 🇩🇰 this week to attend #ICWSM2025. My 4th time attending @icwsm.bsky.social to present our accepted long paper. This time it’s special for me bcz of being in the organizing committee as a tutorial co-chair. Looking forward to meeting you all at #ICWSM25.
Will the meta-review be visible to authors? If so, when?
Ph.D. Hooding Ceremony | May 15, 2025 👩🎓🎓Grateful to the Department of Computer Science at Purdue University & my advisor Dr. Dan Goldwasser for their support. Endless thanks to my partner Masud❤️,my amazing parents💕, & my greatest joy (my kids) Teva & Tivon🩷💙Lucky to have my parents at commencement🎉