Probabilistically split concatenated words using NLP based on English Wikipedia unigram frequencies. - View it on GitHub