So the only words you want to recognize are
„Foods, water, go, come, call me“ ?
That makes it easier 
Are these words examples, or are they a real application for someone who wants food and water?
In this case, a glove and a text to speech software is somewhat overkill.
Give the person 5 buttons, which are labelled accordingly.
From the above paragraph, you can see that you really need to tell us what you are intending to do in the end. Talking about bluetooth, gesture recognition and apps is pointless at this stage of your project. It may become relevant