Abstract:
The purpose of this thesis is introduce a way to make a reliable structure for Persian text in the web by finding semantic relations between Persian words.
There is a corpus for Persian text with label for each word (label such as : Noun or Adjective) , There is a bunch of patterns to extraction the relations , system make a comparison and give the relations as a result.(relations such as this words are Ezafe takhsisi or they are Ezafe jensi )
