Latest Breaking News

... when parsing out a natural language query, is to break the sentence down into "tokens" then search its data base for matching patterns to the tokens.

So the phrase: "...a 15 year old hip replacement" gets transformed to the two tokens:
"A 15 year old" and "hip replacement"

It is programmed to re-state the question it's responding to in the format it had been internally tokenized, which hopefully will make clear to the user the specific and unambiguous question it's is answering.

When you compose a query to ask a search engine, it's helpful to do it in such a way that there is no possible ambiguity when it performs the parsing-out of the token phrases.

So something like: "Is it possible that a person will experience back pain caused by a hip replacement operation which had been performed 15 years prior?"

I hope this helps