Precision at K measures the share of relevant items amongst the top K retrieved documents (without taking into account the position of the item in the list).
It is the ratio of correctly identified relevant items within the total recommended items inside the K-long list. Simply put, it shows how many recommended or retrieved items are genuinely relevant.