Zsolt Katona, Professor

Closed (1) Real or Fake: Managing User Authenticity in Ad Auctions

Applications for Fall 2017 are now closed for this project.

It has be shown by analytical modeling that blocking fake users on user-generated-content platform, such as Facebook, can make them click ads. Subsequently, ad revenue decreases as the number of fake clicks increase. The research will use real Facebook pages created by the team to empirically examine fake activities and validate the theoretical results. Specifically, we will come up with predictive metrics to quantify to what extent fake activities affect ad revenue.

Students will design and set up Facebook pages in the study and learn how to automate web data collection. Those who wish to sharpen their technical skills will contribute to developing and improving extant web-scraping programs as necessary and conduct subsequent predictive modeling. This project is especially well-suited for apprentices who wish to pursue an academic career in marketing science or an industry career in tech. By participating in the project, apprentices will develop web design, web scraping, and data analysis skills as well as people skills by working collaboratively with other teammates.

Day-to-day supervisor for this project: Xin Chen, Ph.D. candidate

Qualifications: A hard work ethic, superb attention to details, strong intellectual curiosity, and most importantly a fearless spirit to quickly pick up whatever new tool as needed. A minimum of 6-hour average weekly commitment and we will give priorities to those who wish to commit more time in our consideration.

Weekly Hours: to be negotiated