Extracting O*NET Features from the NLx Corpus to Build Public Use Aggregate Labor Market Data