Quill's library of open source NLP algorithms and data sets. Quill is using a data set of 100,000 sentences exported from Wikipedia's featured articles. These articles are highly edited, and we use these trust worthy sentences to generate syntactical patterns. For example, by stripping parts of speech out of 50% of the sentences, we can compare a library of 50,000 sentence fragments to 50,000 complete sentences. - View it on GitHub