To have excellent AI and NLP, contextual analysis is required for the understanding of text including where part of the text comes from. There are words or even paragraphs that can have different meanings according to the sentence, also that meaning gives the context for that portion of text.  Increasing the understanding of NLU with Contextual Analysis is paramount for these workloads.

Syntactic and sematic are the general structure of a language, dictating the rules of the construction of the sentence, also the meaning of the words themselves.

When you are analyzing text, you start with semantics, but you want to ensure that a word that has different meaning is being understood as we expect in the context of the tex. Here also comes in Anaphor Resolution, how the sentences are related to each other, and understanding what is going on in a phrase or paragraph.

NLP uses two methods for the resolution of the words, one being the phonetic of each word looking at the all metadata around the sentence, and the second is phonology, the word root giving you the grammar context.