Privacy Hacktivism: You can change the future of the Internet

Description

Recent events have shown that people’s privacy and liberty to interact with others online freely without any danger of being monitored is in grave danger. This is where technical people such as yourself are required: You can stem the flow of these attacks on privacy. Armed with a combination of your technical knowledge and a desire for shaping the future of the web, you can join our project, CloudSpaces, which aims to do exactly that. CloudSpaces is a joint industry-academia project with partners including NEC, EPFL, URV, eyeOS, Tissat and Eurecom.

Below we list some of the potential projects that you can be a part of, some of which will appeal to the more programming oriented ones among you, while others will appeal to those who wish to delve into theoretical research, while yet others are a good mixture of both theory and practice. In all these projects, you will be working in close concert with the both of us.

We are developing PrivySeal, an application that has already seen extensive programming effort, with solid theoretical foundations. PrivySeal is a software that provides easily usable privacy technologies to end users who want to share their data on any cloud service. It is based on published research and employs crowdsourcing and a psychologically grounded model for privacy risk estimation. PrivySeal provides three key services: 1) It determines the risk of sharing data items on the cloud; 2) it provides recommendations for risk mitigation and finally 3) it provides privacy enhancing technologies to mitigate the risk (Check our poster here!)

Under PrivySeal, we have the following projects available:

a) Risk Communication Models: Currently, PrivySeal indicates a risk value to users using a value on our Risk Meter. However, the end user remains oblivious of what this value means. We aim to develop Risk Communication models that seek to better inform end-users of the reasons a certain file-sharing operation is deemed risky. This project will involve both theoretical work (for the models) and programming work, for implementing the models.

b) Recommendation of Policies: We have developed recommendation models to provide users suitable policies to apply on their data (e.g., encryption, hiding metadata etc). We seek those interested in development to implement these models. You will be working very closely with us and together we shall figure out methods to translate our theoretical models into deployable code. Thus you will also be able to whet your appetite for research if you so desire.

c) Efficient processing of batch sharing: Currently, PrivySeal lets the user share one file at a time, while showing the risk of sharing that file, and also allowing the user to apply suitable policies on that file (such as encryption etc). However, in practice, often users want to share multiple files. While this is easy enough, it is not clear how PrivySeal should (or could) show the risk attached with multiple files (imagine a batch of 100 files!) and also how the user would be able to apply policies on them individually. Some intuitive ideas present themselves such as showing a risk histogram for the entire batch of files and choosing the same policy for every file. However, these ideas need to be fleshed out in order to better handle batches of multiple files.

d) Developing PrivySeal for Smartphones: Currently, PrivySeal is a desktop based application. We want to create a smartphone version of PrivySeal, a move which will provide it with a larger user base. We seek someone who has some knowledge of the Android environment. This will be a noteworthy and stimulating programming experience for anyone who takes up this challenge.

Apart from PrivySeal, we also have other potential project ideas for online privacy and would be happy to discuss them.

Prerequisites

• Experience in Java Programming
• For those who wish to develop for the smartphone, some knowledge of the Android environment

Bonus

• Helpful but not essential is a passion for interdisciplinary work.
• Some knowledge of machine learning techniques would be a plus
• A passion for doing work that makes a difference

Contacts

In case of any questions, please drop us an email or come to our offices:

• Rameez Rahman (BC 142)
• Hamza Harkous (BC 168)