This repository contains the data and python scripts for running the evaluations found in our IWCS 2015 and ACL 2015 papers (bib below).
The model of reference resolution is called "Words as Classifiers" for reference resolution (WACrr).
Prerequisites:
python2.7
numpy
scipy
nltk (stem.snowmall.GermanStemmer)
sqlite3
sklearn (linear_model, preprocessing.StandardScaler)
For IWCS:
cd IWCS
python wac_rr_IWCS.py
Note: this will only perform the evaluation using speech; it will not interpolate with gaze and gesture
For ACL:
cd ACL
python wac_rr_ACL.py
@InProceedings{kennington-dia-schlangen:2015:IWCS2015,
author = {Kennington, Casey and Dia, Livia and Schlangen, David},
title = {A Discriminative Model for Perceptually-Grounded Incremental Reference Resolution},
booktitle = {Proceedings of the 11th International Conference on Computational Semantics},
month = {April},
year = {2015},
address = {London, UK},
publisher = {Association for Computational Linguistics},
pages = {195--205},
url = {http://www.aclweb.org/anthology/W15-0124}
}
@InProceedings{kennington-schlangen:2015:ACL-IJCNLP,
author = {Kennington, Casey and Schlangen, David},
title = {Simple Learning and Compositional Application of Perceptually Grounded Word Meanings for Incremental Reference Resolution},
booktitle = {Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing (Volume 1: Long Papers)},
month = {July},
year = {2015},
address = {Beijing, China},
publisher = {Association for Computational Linguistics},
pages = {292--301},
url = {http://www.aclweb.org/anthology/P15-1029}
}