Use a supplied working baseline (java based) program as a framework and build a module to work in conjunction with the supplied tool. You will Select an audio/speech tool that is open source and can run stand alone with our tool set.(python may have one). The new module will import folders of pre trained audio file (placed in positive or negative folders) The audio to text engine will then translate the files to text, then save it into the proper folders as txt files. We will supply you 10 audio files 5 pos 5 negative (I will classify them ahead and put them into folders). V5 will then classify run the 3 files through the text clasification tool and classify according to the naive bayes classifier we have in the baseline tool.