This method uses Mutual Information(MI) to find possible words.
Left and Right Branch Entropy to check word boundaries.
This is a open-source-version, you can also improve the speed of calculation by optimizing the data structure.
Python version and something may you need to install:
python 3.6
tqdm
numpy