Feel free to get in touch by posting here in the forums which are monitored continually by the Hackathon team. Or, if email is more your thing, we got that covered too - just email hackathon@mongodb.com
We were fortunate to receive some further links and resources are GDELT that will help all participants to understand and get familiar with this fantastic Dataset -
The Global Knowledge Graph (GKG) is probably the most relevant, since it is about general purpose extraction of core metadata. You can drop the GCAM field, since it doubles the size of the data and while it contains incredibly rich and deep sentiment data is likely overkill for this kind of analysis:
There’s also the entity dataset that uses the NLP API to annotate a subset of articles each day:
You can see how this can be used for detection and contextualization:
And the geographic graph that could showcase geographic analysis in MongoDB:
There are also image annotations:
Then there is the video annotation dataset of television news:
This can be used for interesting things like tracking tweets on TV news:
There is also a rich embedded metadata graph that compiles JSON-LD and META blocks:
There is an excerpted version at:
We hope the above links and details will inspire you to dig deeper into the GDELT resources. Remember, we will be running GDELT sessions at least twice a week on our livestream - do join us.