Shadi Rezapour
banner
rezapour.bsky.social
Shadi Rezapour
@rezapour.bsky.social
Assistant prof. of information science at Drexel CCI.
computational social science, NLP for social good, narratives, stigma, #NLProc #NLP #NLP4SG
Previously at UIUC, Spotify.
she/her. shadirezapour.com
🎓 Plus, huge congrats to @laylab.bsky.social for presenting her thesis proposal at the Websci Doctoral Consortium!
Feeling super proud of our team, especially our brilliant undergrads and PhD students who led, published, and presented such impactful work.
May 22, 2025 at 8:02 PM
🧒 Youth Online Safety in News w/@afsanehrazi.bsky.social,@elhamagk.bsky.social , @md.ekstrandom.net , and Matt Namvarpour,
📰 3.9K+ articles (2002–2024)
📈 Focus shifted to tech platforms & bullying
💬 Sentiment more negative over time
📌 Trends linked to policy & events
👉 dl.acm.org/doi/10.1145/...
May 22, 2025 at 8:02 PM
🤖 Deception: Human vs. AI w two amazing STAR undergrads Minh Trinh & Samiha Zarin🌟
📚 GPT-3.5/4o = more formal, polished lies
🙋‍♀️ Humans = more social, casual deception
⚠️ LLMs still carry subtle biases
A cool dive into the language of deceit
👉 More: dl.acm.org/doi/10.1145/...
May 22, 2025 at 8:02 PM
🎬 Impact of Science Documentaries w/ @ariapessian.bsky.social and collaborators:
📊 New taxonomy for measuring impact
📝 Annotated Amazon reviews
🤖 ML + LLMs used to analyze public sentiment
🌍 Science storytelling = powerful!
👉 More: dl.acm.org/doi/10.1145/...
Won Best Poster Award #websci2025 🏆
May 22, 2025 at 8:02 PM
📘 Stance on Affirmative Action (AA) via Reddit with @ariapessian.bsky.social
🔍 Analyzed Reddit posts on AA
🧠 LLMs built nuanced stance labels
⚖️ Moral values like Fairness shaped narratives
🗣️ Most users were skeptical of current AA policies
👉 More: dl.acm.org/doi/10.1145/...
May 22, 2025 at 8:02 PM
Don’t forget noise canceling headphones 🎧
December 30, 2024 at 8:56 PM
Congrats!
November 27, 2024 at 8:13 PM