0

I was wondering of anyone have a reference of information talking about semantic distance/similarity between 2 words in a sentence BASED on counting the words that separate between these 2 words? Example:

"The student went to the university to learn about the facts of life"
so if the distance threshold is 3 token then:
"student" and "university" will be connected by an edge (distance 3 words)
" facts and "life" will be connected by an edge (distance 1)
"student" and "Fact" will not be connected (distance 8) 
etc.
Nirke
  • 173
  • 1
  • 5
  • 12

1 Answers1

0

To find a semantic similarity between two words, the most simple algorithm would be to use a word lesk algorithm (https://en.wikipedia.org/wiki/Lesk_algorithm), using wordnet. I doubt if there is any method of finding semantic similarity by the distance between the words.

Arun Jayapal
  • 457
  • 1
  • 4
  • 12