We have performed a couple of AI-centric projects where the core functionality has been two fold; collecting information on the internet and detecting sentiment as well as content.
This makes it possible to not only collect information but also decipher what it means and how we should act on it. Is this review positive or negative or, as most often, both? How do we correctly interpret metaphors and sarcasm? This is what we have done and we feel it will be the underlying platform for many projects to come.
One example is to able to correctly categorise a review, pick out what was actually positive or/and negative, rank how much, and compose a relevant reply automatically. Another example is to gather information from the world around and build an opinion om where it is going by also weighing in the sentiment.
Lämna ett svar