This can be a interview question for even advanced linux users.

 

Here is the command : grep -o spellword file | grep wc -w 

 

( where spellword = name of the word you want to count and file is the name of the file )

 

Result : You will find the result based on occurrence of specific word.