Rezvaneh "Shadi" Rezapour
School of Information Sciences (iSchool @ Illinois)
I am fifth year Ph.D. student at the School of Information Sciences, University of Illinois at Urbana-Champaign. I am working as a research assistant in the Social Computing Lab under Professor Jana Diesner's supervision.
My current field of study falls within the domain of computational social science. In particular, I develop models to extract meaningful information from (online) social discourse. More broadly, I am interested in combining methods from natural language processing, machine learning, and network analysis with social science theories to better understand real world behaviors, attitudes and cultures.
My line of research can be divided into two parts:
(1) Impact Assessment: In the past couple of years, I have been a part of "Impact Assessment Project" in Dr. Diesner's lab to study and analyze user-generated texts (e.g.: reviews, grant reports) to extract various types of impacts. We leverage both quantitative (natural language processing and machine learning) and qualitative (close-reading, annotation) models to build and analyze the input data and extract meaningful information from raw texts. In our paper "Classification and Detection of Micro-Level Impact of Issue-Focused Documentary Films based on Reviews", published in CSCW 2017, we analyzed Amazon reviews and found that films can change people's behavior, cognition, and emotion. You can read more about it here.
(2) Studying Personal Biases and Values: We believe that people's behavior, culture, and background impact their personal values which then manifest in people’s social discourse and everyday use of language. Morality is one of the aspect that can portrait individual values. We believe that morality can enhance studying and predicting social effects such as stance, emotion and sentiment. We have recently developed and published an enhanced version of the moral foundations dictionary. In our recent paper published in the WASSA workshop at NAACL, we showed that both classic machine learning and deep learning models benefit from morality as an NLP feature. Read more about it here.
Most Recent News
- December 2019: Organizng the 2nd Workshop on Computational Impact Detection from Text Data @ LREC 2020; due: February 20th, 2020
- December 2019: Organizing the first student poster session at WIDS UIUC , March 6 2020, NCSA
- November 2019: I will be atteding CSCW to present our work on morality and social movements.
- October 2019: I will attend GHC to present a poster at ACM SRC.
- August 2019: Our extended abstract on social movements and morality is accepted in CSCW 2019.
- June 2019: Going to NAACL 2019 to present our paper on morality and social effects.
- May 2019: My poster is accepted at GHC 2019 for the ACM Student Research Competition.
- April 2019: I am going to present our paper "Enhancing the measurement of social effects by capturing morality" at the Midwest Speech and Language Days.
- March 2019: Our work "Enhancing the measurement of social effects by capturing morality" is accepted in WASSA workshop co-located with NAACL 2019.
- March 2019: Our abstract has been accepted for the 2019 Sunbelt Conference in Montréal.