Professor John Barrios is seeking students to assist with his research that while grounded in economics, requires a skillset in computer science, using natural language processing to generate measures of information from texts in social media sites and newspapers.
Professor Barrios's research focuses on 1) the relation between the governance of police departments and policing behavior in cities, and 2) the role of private incentives on social media activities of academics. To conduct the research, he is compiling several novel datasets that combine information on police departments, police union contracts, police traffic stops, etc.
The student would support the research project with the following tasks:
- Collect data on police behavior from various public sources – requires some scraping of websites
- Use publicly available APIs to infer information about analysts’ ethnicity and gender-based on their names
- Use publicly available python packages to perform natural language processing determining the tone and sentiment of tweets for example
- Merging various datasets using STATA
- Prepare graphs and descriptive statistics that describe the characteristics of the sample of analysts (50 hours)
- Perform OLS regression analysis using STATA
- R, STATA programming, Latex
- Prior coursework in statistics
- Prior coursework in Applied Regression Analysis, natural language processing, and computer programming (preferably CMSC courses on database management, python and functional programming)
Intererested students should contact Professor John Barros at firstname.lastname@example.org.