1

in my experiment, I am using WordNet18 data, which contains triplets of the form (subject, predicate, object), some examples are as follows:

03964744 _hyponym 04371774 00260881 _hypernym 00260622 02199712 _member_holonym 02188065

I would like to know what the entity IDs, like 03964744, stand for. Is there anybody knows how to decode the entities?

Thank you in advance. 

Roger
  • 161
  • 1
  • 9

1 Answers1

1

The 8-digit numbers you see are probably the byte offset of the entry in the data files. See http://wordnet.princeton.edu/wordnet/man/wnintro.5WN.html

After quite a bit of hunting around, I think you are looking at the numbers of WordNet 3.0 (the byte offsets for a given synset differ between versions; 3.1 is the latest version).

Your first entry seems to be saying that swing is a type of toy:

http://compling.hss.ntu.edu.sg/omw/cgi-bin/wn-gridx.cgi?usrname=&gridmode=grid&synset=04371774-n&lang=eng&lang2=eng

http://compling.hss.ntu.edu.sg/omw/cgi-bin/wn-gridx.cgi?synset=03964744-n

Darren Cook
  • 27,837
  • 13
  • 117
  • 217