I have a pool of knowledge that I want to mine for information and allow an AI to deduce likely conclusions from this information.
My goal is to give the AI a set of textual data that is rated on a scale of 0 to 100 ranging from false (0) to unequivocally true (100). Based on ongoing learning I want to be able to ask it about it's data and to make relational conclusions as to not simply whether things are true or false, but to extrapolate likelihoods, conclusions and so forth... or to simply tell me it can't understand something which would then trigger me to give it more information and to train it with additional material - even if it's my own limited answers.
Ultimately I'll deal with image data as well, but that's a bit down the road.
I'm new to the area of neural nets and deep learning and so I'm hoping someone could point me in the right direction in the way of terminology to search for / research as well as where perhaps I should start.
I wouldn't mind working in C predominantly if possible, but other languages (especially Ruby) is fine.
The field is moving so fast and there's so much research now that seems to trump information available from just a couple years ago now and so I'm hoping to jump into information that takes advantage of more general learning algorithms so that this can be as robust possible while taking care of current trends.
Where do I go from here?
