Research Projects

Detecting expressive writing in online health communities by modeling aggregated data

Conducted a meta-analysis to collect aggregated linguistic features of expressive writing

Built a Gaussian naive Bayes model based on meta-analysis results to detect expressive writing

Developed a codebook to label expressive writing in the wild and tested the performance of the classifier

Applied logistic regression to study what factors affected expressive writing</br>

Building a conceptual framework to organize the literature of affective computing

Reviewed over 2000 papers searched from ACM and IEEE digital libraries to filter out irrelevant ones

Proposed a theoretical framework (function-component-representation) to organize the literature

Coded 1110 papers based on the framework

Summarized research progress and discussed insights and future directions</br>

Understanding how expressive writing and social support influence user engagement

Developed classifiers to detect deceased sites which ended due to patients’ literal deaths

Applied survival regression to study how expressive writing and social support influenced user engagement</br>

Examining effects of anonymity, ephemerality, and system routing on cost in social Q&A

Built a social Q&A web platform using Ruby on Rails

Conducted experiments to test effects of anonymity, ephemerality, and system routing on social costs

Applied linear regression models to analyze the user data