About

I am a research scientist in the Responsible AI team at Google Research. My research is dedicated to advancing the development and evaluation of NLP models by incorporating social, psychological, and cultural factors that shape our perceptions of language. Focusing on established social scientific theories and NLP models, my research aims to make language models more representative of the complexities and distinctions of humans and cultures.

Before joining Google Research, I was a research assistant in the Computational Social Science lab at University of Southern California (USC), where I received my PhD in Computer Science working on the impact of cognitive biases and stereotypes on the design and implementation of supervised NLP models. My research has led to NLP models for incorporating annotators' perspectives into text classifiers, identifying moral expressions in language, detecting unreported hate crime incidents from local news articles, and mitigating bias in hate speech classification.

Feel free to Email me, check my CV, explore my Google Scholar, clone my projects on GitHub, or even connect via LinkedIn!

Publications

• Hate speech classifiers learn normative social stereotypes
  Aida Mostafazadeh Davani, Mohammad Atari, Brendan Kennedy, Morteza Dehghani
  TACL, 2022.
• Pathogens are linked to human moral systems across time and space
  Mohammad Atari, Joseph Hoover, Brendan Kennedy, Aida Mostafazadeh Davani, Ali Omrani, Farzan Karimi-Malekabadi, Gwenyth Portillo-Wightman, Shirin Birjandi, Morteza Dehghani
  Current Research in Ecological and Social Psychology, 2022.
• Dealing with disagreements: Looking beyond the majority vote in subjective annotations
  Aida Mostafazadeh Davani, Mark Díaz, Vinodkumar Prabhakaran
  TACL, 2021.
• On releasing annotator-level labels and information in datasets
  Vinodkumar Prabhakaran*, Aida Mostafazadeh Davani*, Mark Díaz
  LAW-DMR Workshop @ EMNLP, 2021.
• Improving counterfactual generation for fair hate speech detection
  Aida Mostafazadeh Davani, Ali Omrani, Mohammad Atari, Brendan Kennedy, Xiang Ren, Morteza Dehghani
  WOAH Workshop @ ACL, 2021.
• Morally homogeneous networks and radicalism
  Mohammad Atari, Aida Mostafazadeh Davani, Drew Kogon, Brendan Kennedy, Nripsuta Ani Saxena, Ian Anderson, Morteza Dehghani
  Social Psychological and Personality Science, 2021.
• The gab hate corpus: A collection of 27k posts annotated for hate speech
  Brendan Kennedy, Mohammad Atari, Aida Mostafazadeh Davani, Leigh Yeh, Ali Omrani, Joseph Hoover, ..., Morteza Dehghani
  Language Resources and Evaluation, 2021.
• Investigating the role of group-based morality in extreme behavioral expressions of prejudice
  Joseph Hoover, Mohammad Atari*, Aida Mostafazadeh Davani*, Brendan Kennedy*, Gwenyth Portillo-Wightman, Leigh Yeh, Morteza Dehghani
  Nature Communications, 2021.
• Moral concerns are differentially observable in language
  Brendan Kennedy, Mohammad Atari, Aida Mostafazadeh Davani, Joseph Hoover, Ali Omrani, Jesse Graham, Morteza Dehghani
  Cognition, 2021.
• On transferability of bias mitigation effects in language model fine-tuning
  Xisen Jin, Francesco Barbieri, Brendan Kennedy, Aida Mostafazadeh Davani, Leonardo Neves, Xiang Ren
  NAACL, 2021.
• Hatred is in the eye of the annotator: Hate speech classifiers learn human-like social stereotypes
  Aida Mostafazadeh Davani, Mohammad Atari, Brendan Kennedy, Shreya Havaldar, Morteza Dehghani
  CogSci, 2020.
• Contextualizing hate speech classifiers with post-hoc explanation
  Brendan Kennedy*, Xisen Jin*, Aida Mostafazadeh Davani, Morteza Dehghani, Xiang Ren
  ACL, 2020.
• Moral Foundations Twitter Corpus: A collection of 35k tweets annotated for moral sentiment
  Joe Hoover, Gwenyth Portillo-Wightman, Leigh Yeh, Shreya Havaldar, Aida Mostafazadeh Davani , Ying Lin, Brendan Kennedy, Mohammad Atari, Zahra Kamel, Madelyn Mendlen, Gabriela Moreno, Christina Park, Tingyee E Chang, Jenna Chin, Christian Leong, Jun Yen Leung, Arineh Mirinjian, Morteza Dehghani.
  Social Psychological and Personality Science, 2020.
• Reporting the unreported: Event extraction for analyzing the local representation of hate crimes
  Aida Mostafazadeh Davani, Leigh Yeh, Mohammad Atari, Brendan Kennedy, Gwenyth Portillo-Wightman, , Natalie Delong, Rhea Bhatia, Arineh Mirinjian, Xiang Ren, Morteza Dehghani
  EMNLP, 2019.
• Modeling performance differences on cognitive tests using LSTMs and skip-thought vectors trained on reported media consumption
  Maury Courtland, Aida Mostafazadeh Davani, Melissa Reyes, Leigh Yeh, Jun Leung, Brendan Kennedy, Morteza Dehghani, Jason Zevin
  NLP+CSS Workshop @ NAACL, 2019.
• Subtle differences in language experience moderate performance on language-based cognitive tests.
  Maury Courtland, Aida Mostafazadeh Davani, Melissa Reyes, Leigh Yeh, Jun Leung, Brendan Kennedy, Morteza Dehghani, Jason Zevin
  CogSci, 2019.
• Body maps of moral concerns
  Mohammad Atari, Aida Mostafazadeh Davani, Morteza Dehghani
  Psychological Science, 2019.

Pre-prints and In-progress

• Syntactic and semantic gender biases in the language on children’s television: Evidence from a corpus of 95 shows from 1960 to 2018
  Andrea C. Vial*, Aida Mostafazadeh Davani*, Shreya Havaldar, Eleanor K. Chestnut, Morteza Dehghani, Andrei Cimpian
  
• Twitter observers’ moral language reveals how sexual harassment denials condemn #MeToo victims
  Rachael D. Goodwin, Samantha J. Dodson, Michelle Chambers, Aida Mostafazadeh Davani, Morteza Dehghani, Jesse Graham, Kristina A. Diekmann