colinsongf/retrofitting

Retrofitting Word Vectors to Semantic Lexicons

#Retrofitting Manaal Faruqui, [email protected]

This tool is used to post-process word vectors to incorporate knowledge from semantic lexicons. As shown in Faruqui et al, 2014 these word vectors are generally better in performance on semantic tasks than the original word vectors. This tool can be used for word vectors obtained from any vector training model.

###Requirements

1. Python 2.7

###Data you need

1. Word vector file
2. Lexicon file (provided here)

Each vector file should have one word vector per line as follows (space delimited):-

the -1.0 2.4 -0.3 ...

###Running the program

python retrofit.py -i word_vec_file -l lexicon_file -n num_iter -o out_vec_file

python retrofit.py -i sample_vec.txt -l lexicons/ppdb-xl.txt -n 10 -o out_vec.txt

where, 'n' is an integer which specifies the number of iterations for which the optimization is to be performed. Usually n = 10 gives reasonable results.

###Output File: out_vec.txt

which are your new retrofitted and (hopefully) improved word vectors, enjoy !

###Reference

Main paper to be cited

@InProceedings{faruqui:2014:NIPS-DLRLW,
  author    = {Faruqui, Manaal and Dodge, Jesse and Jauhar, Sujay K.  and  Dyer, Chris and Hovy, Eduard and Smith, Noah A.},
  title     = {Retrofitting Word Vectors to Semantic Lexicons},
  booktitle = {Proceedings of NAACL},
  year      = {2015},
}

If you are using PPDB (Ganitkevitch et al, 2013), WordNet (Miller, 1995) or FrameNet (Baker et al, 1998) for enrichment please cite the corresponding papers.