Is GDELT’s HR-GKG Useful for Human Rights Researchers?

Back in Sept of 2014 this happened on twitter:

Moore & Pruce Twitter Exchange

Moore & Pruce Twitter Exchange

In January I explored GDELT’s HR-GKG a bit (that brief paper can be found here [pdf]), and today I am finally writing a brief post about it.  Here is how I conclude said paper:

Unfortunately I have not been successful locating promising opportunities for using the HR-GKG data to augment or extend the ITT SA data. Nor have I located opportunities for using it in other ways. It is important to observe that this does not suggest that the HR-GKG is flawed or “useless.” To make such judgments one would need to know the specific purposes the GDELT project had in mind when designing and executing the project, and I have not been able to locate documentation of that…

That said, I am hard pressed to come up with ways for academic researchers broadly interested in the types of questions I have been studying in recent years to put the HR-GKG to good use. A handful of shortcomings with respect to such efforts stand in the way. First, the absence of dates that can be tied to events or themes is a likely deal breaking impediment to the utility of the HR-GKG for academic researchers interested in the naming and shaming activities of organizations.

Second, it is not clear what are the possible values for the Counts field, and that makes it challenging to assess the usefulness of the data for any given application. When one turns to produce frequencies to ascertain the set of values one encounters the third shortcoming: the data are not structured in a conventional format that most academic researchers who work with country–year or events data will find familiar. Since the HR-GKG was not, to my knowledge, produced with academic researchers in mind, this is not a criticism, but rather a note to academic researchers who wish to use the data. Recognize going in that there will be data management issues that may well fall outside of your experience and skill set. Relatedly, the absence of descriptives limits researchers’s ability to gauge the likely value of the data for their specific project absent addressing the data reshaping.

If you decide that you would like to explore that HR-GKG data, you may find these R scripts  useful for getting you started (thanks for Chris Fariss for an assist with them).


About Will H. Moore

I am a political science professor who also contributes to Political Violence @ a Glance and sometimes to Mobilizing Ideas . Twitter: @WilHMoo
This entry was posted in Uncategorized and tagged , , . Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s