Welcome to tux02ascor.fmg.uva.nl!

This server provides you with all you need to conduct large-scale and real-time media analysis (hence the project's name, LSRTMA). We do our best to make this as accessible as possible.

What do you need?

(Level 1) I just want an online search interface to the media data on this server

You probably want to use our AMCAT4 interface for that. [SORRY, NOT FULLY OPERATIONAL YET]

(Level 2) I do not have much coding experience, but maybe you can show me how to do [analysis X]?

Yes we can! We prepared several use cases for you, with detailed instructions on how to run your analysis. Think of answering questions like:

How often is organization mentioned in the news?
How does the number of articles containing phrase X develop over time>
Is there a linguistic difference between news coverage in outlet A and outlet B?
...

(Level 3) I know what I'm doing, I just want access

You can make use of this server through JupyterLab, which allows you to run code in either Python, R, or Julia. In order to make use of it, you only need to provide us with your Github username. After we have verified that your request is legitimate and have added your Github username to our system, you can log on to Jupyterhub via Github via https://tux02ascor.fmg.uva.nl/jupyter. Note that the service is only avaible from within the UvA network (use a VPN when at home). Logging on will spawn a Jupyter instance in a Docker container. You can imagine this as a little virtual machine only for you. What you run in your instance does not effect others (except, of course, that everything gets slower for everyone if you use too many ressources, and that others get into trouble if you use too much storage). During your first visit, you will find the following files in your environment:

welcome.ipynb - A notebook that introduces you to working with the server
shared-ro/ - A folder for shared content for all users. It's read-only
shared-rw/ - A folder for sharing content with all other users. It's read-write, so you can write to it. It can be useful for data exchange. Keep in mind that everbody can read from it, so do not write sensitive data on here. There is also no guarantee that content persists - we will clean up regularly.

We already pre-installed most data-scieny packages, but because you are running your own Docker container, you can install additional packages yourself. Just run the following code to install the package called df2markov (pay attention to the exclamation mark):

!pip install df2markov

(Level 4) I need more!

On request, it is possible to grant you SSH access and the right to spawn docker containers. In particular, this means that you can run your own database (e.g., ElasticSearch) as a backend and access it through Jupyter. For instance, you could run your own INCA or AMCAT.

Code of conduct

This is a shared space. We want to make using this environment a nice experience for everyone. Therefore, by using this environment, you commit to the following code of conduct:

Be nice! Do not use more ressources than you really need.
Clean up and remove data when you do not need them any more
The service is only to be used for the intended purpose. In particular, any use that is not for educational and research purposes is explicitly forbidden.

No guarantees!

This service is provided on a best-effort basis. Using it is at your own risk. In particular, it is your own responsibility to arrange long-term storage of your data. While we do backup the server regularly, we cannot guarantee that your data will be kept (in particular if your project is finished and/or if you use excessive amounts of space). Be aware that -- depending on how the permissions of your files and folders are set, and where you store the data -- it cannot be fully guaranteed that others can read them. If you have particularly sensitive data to process, discuss this first.

Good luck with your work!