Undergrad Research Project - HCII PrivacyGrade Project

Spring 2017

Student
Alexander Yu
Advisor
Jason Hong
Project description

The PrivacyGrade project in the HCII department takes Android smartphone apps in the Google PlayStore and assigns each app a grade based on how good a job it does with data privacy. The project is meant to show smartphone users what kind of data and levels of access they are agreeing to give to these apps. The project consists of 3 main parts; the crawler that trawls through all the apps in the PlayStore and downloads their apk files, the analysis pipeline that analyzes all the apps to ascertain their “privacy grade”, and the website for the general public to browse our results with.

Researchers in a wide range of fields are analyzing smartphone apps, examining issues of privacy, security, usability, accessibility, and more. However, many of these analyses only examine a few hundred or a few thousand out of the million-plus apps that are available. Furthermore, while there have been many papers analyzing smartphone apps, very little software or data analyses have been shared, making it difficult for researchers to build on top of each other’s work.

To address the above problems, we will build a community platform to lower barriers to analysis and sharing by developing and hosting an analysis platform that will let researchers upload scripts to analyze apps and share their work in a simple manner. This will also offer a hub for researchers to allow them to see new kinds of analyses, share data and materials, and discuss tools and techniques for analyzing apps. We will develop new kinds of workflow tools to make it easy for researchers and other interested parties (such as journalists and government agencies) to explore the data and/or pose research questions for the community, thereby helping the matchmaking between needs and research. In the future, we also plan offer tutorials and scaffolding tools to enable students to contribute, helping them learn about app analysis while also offering advanced students opportunities for scaling up their work.

The goal of this project is to build the basics of a community platform that greatly facilitates the task of analyzing smartphone apps at scale, building useful predictive models, and quickly sharing the results with the community and the public. The basics will allow people to utilize our data, or certain parts of it, remotely, upload code that accesses and operates on this data, and then receive the results. The next addition would be having basic security checks to make sure the code does not compromise our data in anyway. The goal for the semester is to have this up and running for internal usage, which will allow our own developers to easily run analyses.

Return to project list