Intuition is massive parallel pattern matching. However, it’s not enough to match patterns, a system must be able to (as you allude to) separate the important from the unimportant. How does it perform that separation, in learns that from ‘experience’ and from ‘imagination’. So the next time its sees a complex situation, it is able to (using massively parallel pattern matching) find what is relevant, unusual and eventually what is most important.
This is intuition and humans use this all the time to understand the world. This is how Deep Learning machines work to recognize images or voices.