Updated the description of evaluation results, added info about the corpus-based error-correction list.

Minor change in the 'posgram_errors' method output.

Added the part-of-speech n-gram based error finder to the repository.

Update config.py