Nice, thanks! I wish there was a cleaned up repository of datasets like these, in a unified format, directly accessible to a public MapReduce engine like Elastic at AWS.