Is this the same idea as CMU's NELL?

Here is a simple, broad idea towards how to build commonsense knowledge bases using Linguistics. Take a general sentence – Doctor is doing surgery on a patient. Now keep expanding the meaning of every word in the sentence (and the words in it, and so on...) using a dictionary and wikipedia, and keep attaching the ‘discoveries’ in each such process, to the agents in the original sentence. So with every random ‘general sentence’ you automatically get a few commonsense facts fed into the computer. In the above example - The expansion of ‘patient’ will unveil the aspect of “medical treatment” which further expanded would bring in the concept of “cost/paymeny/charges/fees to the doctor”. This would create the commonsense fact (by associating these discoveries to the original sentence) – When doctors do surgeries (/medical treatment) on patient, the latter pays them fees/money. The expansion of ‘surgery’ will unveil the aspect of “cutting the skin (mostly)” which would get attached to the orignial sentence as a commonsense fact – When doctors do surgery on a patient, they cut open the skin mostly.