I was reading the recent paper https://www.aclweb.org/anthology/P19-1580.pdf and noticed that in section 5.2, the syntactic relation is studied in terms of the "direction between two tokens". In table 1, the result is further shown with direction like $$v \to s, s \to v$$ for nsubj(nominal subject).

enter image description here

However, what does V and S refer to here and where I can find more materials about this ?


1 Answer 1


This is explained in the Methodology section of the paper:

We calculate for each head how often it assigns its maximum attention weight (excluding EOS) to a token with which it is in one of the aforementioned dependency relations. We count each relation separately and allow the relation to hold in either direction between the two tokens.

The nsubj dependency label is used in arcs between the verb ($v$) and a subject ($s$). In the paper, they define this "dependency score" taking the maximum attention weight and checking if it points to the expected word according to the dependency parse arc, in either of the directions. This is why they count it from verb to subject and from subject to verb.

If you want to know more about the conventions usually followed in dependency parsing, including these nsubj, amod, etc, you can take a look at the Stanford typed dependencies manual.

  • $\begingroup$ Thx, then what is S in $$v \to s$$ referred to here ? $\endgroup$
    – exteral
    Commented Jun 7, 2020 at 21:12
  • $\begingroup$ $s$ is for subject. Sorry, in the answer I had changed "subject" ($s$) with "noun" ($n$). It's fixed now. $\endgroup$
    – noe
    Commented Jun 8, 2020 at 7:06

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.