How does Artificial Intelligence and semantic search work for a portal

This is a very vague question .. and sorry to say I wont be able to answer you in a better way...

For the sake of conversation , lets say you own clarity.fm and see how we can fit semantic and AI on to it...

I had once started a similar idea in past on yelp reviews..

Semantic - In layman langugae, you are breaking down large language pieces to small sentences and making a meaning out of it...and tagging them..
E.g If there was an sematic alogorithm you had run on this particular question , then the results would be
Vagueness = 10, topic = technology, industry = web etc..

TO get this done, there should be a lexicon database that you should seed and curate.

AI - This is more about applying data science and doing doing deep linking, mapping between all the raw data points (like User details) and derived information (like semantics) and identifying patterns. This is just one example.

One practical example using clarity as case study using this clarity question --- If today you want to make a call for some help, calrity can use its AI algorithms to to identify who would be the best person to answer this..(based on the answers made in the past).

This is a vast topic...
To have a better answer , you need to share more info..
Feel free to talk. Not sure where you are based but if you can't afford the $, ask from me for a free call.. more than happy to help for 30 mins.. love the challanging part.