Semester Project

 

This page summarizes some resources to help you to conduct the semester project.

Dataset

There are multiple online datasets.

  • SNAP from Stanford (recommended): http://snap.stanford.edu/data/index.html
  • Arizona State University: http://socialcomputing.asu.edu/pages/datasets
  • UCI dataset: https://archive.ics.uci.edu/ml/datasets.html

Datasets we have at CISPA:

  • Location datasets from Instagram (https://yangzhangalmo.github.io/papers/CCS17.pdf)
  • Hashtag datasets (https://yangzhangalmo.github.io/papers/WWW18.pdf)

You can also collect datasets yourself

  • Twitter's API with Tweepy

Some papers

Attribute Inference:

  • paper: https://cs.stanford.edu/~jure/pubs/node2vec-kdd16.pdf
  • video: https://www.youtube.com/watch?v=1_QH5BEP5BM

Unicity of Location:

  • paper: https://www.nature.com/articles/srep01376
  • video: https://www.youtube.com/watch?v=DPqreIYe0UU
  • video: https://www.youtube.com/watch?v=kLMKYORwjTk

Membership Inference of Machine Learning:

  • paper: http://www.comp.nus.edu.sg/~reza/files/Shokri-SP2017.pdf
  • video: https://www.youtube.com/watch?v=rDm1n2gceJY

Stealing Machine Learning Models:

  • paper: https://www.usenix.org/system/files/conference/usenixsecurity16/sec16_paper_tramer.pdf
  • video: https://www.usenix.org/conference/usenixsecurity16/technical-sessions/presentation/tramer

walk2friends:

  • paper: https://yangzhangalmo.github.io/papers/CCS17.pdf
  • video: https://www.youtube.com/watch?v=GSQ9JPuvU6U

Tagvisor:

  • paper: https://yangzhangalmo.github.io/papers/WWW18.pdf

Programming

It is highly recommended to use Anaconda to do your project, Anaconda is a data science distribution of python, it contains almost all the packages you need for your project.

Anaconda can be downloaded here: https://www.anaconda.com/download

The packages we need are:

pandas, numpy, scipy, and scikit-learn

There are multiple videos on youtube that introduce pandas, you can take any of them and have a quick look,

e.g., https://www.youtube.com/watch?v=-NR-ynQg0YM

 

 

Privacy Policy | Legal Notice
If you encounter technical problems, please contact the administrators.