Datasets - crimson-unicorn/core GitHub Wiki

Use Ours or Use Your Own, Your Choice

We provide several sets of data at your disposal. You can reproduce our results and verify our experiments with those datasets, or you can try our framework using your own data.

If you would like to use our datasets, you have many choices, including StreamSpot and CamFlow datasets. In this tutorial, we use a toy dataset, which is a subset of the StreamSpot datasets. Therefore, you must download StreamSpot first:

make download_streamspot

:copyright: We don't own the rights to StreamSpot (or some other datasets like DARPA TC), but they are all publicly available online.

:floppy_disk: To make things easier and more predictable for us, we have transported some data to DataVerse and some others to GitHub. You can find the original DARPA TC E3 data here and E5 data here.

The data should all be downloaded to a data/ directory under core/ (the command above creates such a directory for you). You are welcome to use your own dataset, but remember, you need to write your own parser so that Unicorn can understand your data. You can learn more about how you can write a parser by understanding the documentation here.

If you haven't clone the code, do it now before running the experiments.