2012-2013 BDSS IGERT Events

Trainees participate in challenges & competitions!


GDELT Hackathon I:

GDELT (Global Event Location & Tone) is a new CAMEO-coded data set containing more than 200-million geolocated events with global coverage for 1979 to the present. The data are based on a variety of international news sources coded using the TABARI system for events and additional software for location and tone and is updated daily.

The Hackathon was held September 16, 2012 at PSU with the IGERT team and affiliates as well as other visiting and PSU faculty and students working through the day with the data set to assess how to work with the data, how to aggregate it, what initial questions could be asked and given subsets of the data, what output could be generated with this initial work. The “results” covered a range from possible ways to handle the data, aggregate it, and then pull out the findings and represent them in a visually interesting manner.image cache



kaggle Competition - Predict census mail return rates

The IGERT team competed in this challenge in the Fall of 2012. The challenge was to create the best statistical model predicting 2010 Census mail return rates at block group level geographies based on their demographic characteristics. The contest attracted 244 entries.

 The IGERT team placed in the top 10% of entries!

IGERT Video/Poster Competition

 The IGERT team entered a video poster in the 2013 IGERT Competition. Josh Stevens, PhD candidate in Geography and a BDSS IGERT trainee provided the leadership for this entry.  The video is accessible here: http://www.youtube.com/watch?v=YW7shbEUrXU

Big Data Week 2013 @PSU Mini-Conference

Big Data Week (bigdataweek.com) is a global, week-long 'conference' on Big Data held across many cities and countries, embracing multiple formats and themes. BDSS IGERT @PSU participated with a series of presentations during the week of April 22nd, 2012. The PSU themes included 'R and Big Data', 'Scaling Social Science Up', 'Managing Geospatial Data' and 'Measuring Latent Concepts'.