This project first seeks to measure the Internet connectivity of users at locations in the United States and use the collected data to create a map of Internet connectivity. Slides from a talk on this topic are available. See GeoConnected Project for map on physical connectivity.
Suggested Background: Familiarity with network protocols such as TCP, HTTP and DNS as well as network and scripting tools are desirable. The project may also involve GeoIP services.
Third-party advertisers track the behavior and learn much information users on the Web and mobile platforms. Concerned users install ad blockers to limit such tracking, but often do so without understanding what these ad blockers are doing. The goal of this project is to create a tool using semantic-based specifications compiled directly from user responses to desired privacy levels. The tool will then generate a customized set of rules for different ad blockers and platforms.
Suggested Background: Familiarity with Webware and Networking.
It is common in sports that teams (or players) are ranked based on head-to-head competition. These head-to-head matches provide direct evidence on which team is better, but converting these "partial orderings" into a total ordering (ranking) of teams can be difficult. This project will gather sports data and use it to explore different approaches to rankings that minimize partial ordering violations. Other factors such as margin of victory, home team adjustment and recency of match will also be considered.
This is an example of the feedback arc set problem, which is NP-complete, but has many approximations. We will be looking to examine how to apply these approximations to this problem and to understand how the results compare with other ranking approaches.
Suggested Background: Familiarity with data analysis, scripting, and algorithms.
This project seeks to gather data on the academic and athletic prominence of American universities. These data will be used to understand the relationship between these two qualities. Other factors such as type/size of school, athletic division/conference and academic reputation may also be considered.
Suggested Background: Interest in statistics and data analytics.