firem 6 years ago

We analyzed 38,000 Networks, 2.5 Million Graphs -- 1 TB of compressed data -- constructed form Reddit, Microsoft Academic, Chess Games (FICS), Bitcoin, and WikiTree to investigate how trends change over time, to understand how networks evolve, and to understand how certain users gain and lose power over time. We did our best to make our research as open and accessible as possible.

If you want more details about our study, you can:

- Watch our short video (https://youtu.be/zZNZ03ULASk)

- Read our paper (https://arxiv.org/abs/1706.06690)

- Play with the D3 based interface to see how different Reddit/Research Field communities change over time (https://dynamics.cs.washington.edu)

- Download the data (https://dynamics.cs.washington.edu/data.html)

- Read our code tutorials (https://dynamics.cs.washington.edu/code.html)

The data can be really helpful to people that are looking for massive networks/temporal network datasets. For most networks, the data also include the time each two users interacted. As part of this study, we also released about 38,000 time-series that can be used to understand how trends change.

If you have any questions about this study, please feel free to contact me