New Experiments in Distributional Representations of Synonymy
D. Freitag, M. Blume, J. Byrnes, E. Chow, S. Kapadia, R. Rohwer, and Z. Wang. Proceedings of the Ninth Conference on Computational Natural Language Learning, page 25--32. Stroudsburg, PA, USA, Association for Computational Linguistics, (2005)
Abstract
Recent work on the problem of detecting synonymy through corpus analysis has used the Test of English as a Foreign Language (TOEFL) as a benchmark. However, this test involves as few as 80 questions, prompting questions regarding the statistical significance of reported results. We overcome this limitation by generating a TOEFL-like test using WordNet, containing thousands of questions and composed only of words occurring with sufficient corpus frequency to support sound distributional comparisons. Experiments with this test lead us to a similarity measure which significantly outperforms the best proposed to date. Analysis suggests that a strength of this measure is its relative robustness against polysemy.
Description
New experiments in distributional representations of synonymy
%0 Conference Paper
%1 freitag2005experiments
%A Freitag, Dayne
%A Blume, Matthias
%A Byrnes, John
%A Chow, Edmond
%A Kapadia, Sadik
%A Rohwer, Richard
%A Wang, Zhiqiang
%B Proceedings of the Ninth Conference on Computational Natural Language Learning
%C Stroudsburg, PA, USA
%D 2005
%I Association for Computational Linguistics
%K evaluation semantics
%P 25--32
%T New Experiments in Distributional Representations of Synonymy
%U http://dl.acm.org/citation.cfm?id=1706543.1706548
%X Recent work on the problem of detecting synonymy through corpus analysis has used the Test of English as a Foreign Language (TOEFL) as a benchmark. However, this test involves as few as 80 questions, prompting questions regarding the statistical significance of reported results. We overcome this limitation by generating a TOEFL-like test using WordNet, containing thousands of questions and composed only of words occurring with sufficient corpus frequency to support sound distributional comparisons. Experiments with this test lead us to a similarity measure which significantly outperforms the best proposed to date. Analysis suggests that a strength of this measure is its relative robustness against polysemy.
@inproceedings{freitag2005experiments,
abstract = {Recent work on the problem of detecting synonymy through corpus analysis has used the Test of English as a Foreign Language (TOEFL) as a benchmark. However, this test involves as few as 80 questions, prompting questions regarding the statistical significance of reported results. We overcome this limitation by generating a TOEFL-like test using WordNet, containing thousands of questions and composed only of words occurring with sufficient corpus frequency to support sound distributional comparisons. Experiments with this test lead us to a similarity measure which significantly outperforms the best proposed to date. Analysis suggests that a strength of this measure is its relative robustness against polysemy.},
acmid = {1706548},
added-at = {2017-12-17T13:01:21.000+0100},
address = {Stroudsburg, PA, USA},
author = {Freitag, Dayne and Blume, Matthias and Byrnes, John and Chow, Edmond and Kapadia, Sadik and Rohwer, Richard and Wang, Zhiqiang},
biburl = {https://www.bibsonomy.org/bibtex/2c715915ca77e2184214723b41ca42374/thoni},
booktitle = {Proceedings of the Ninth Conference on Computational Natural Language Learning},
description = {New experiments in distributional representations of synonymy},
interhash = {28a24388a3ddb1cc31d70d1e7602f3da},
intrahash = {c715915ca77e2184214723b41ca42374},
keywords = {evaluation semantics},
location = {Ann Arbor, Michigan},
numpages = {8},
pages = {25--32},
publisher = {Association for Computational Linguistics},
series = {CONLL '05},
timestamp = {2017-12-17T13:01:21.000+0100},
title = {New Experiments in Distributional Representations of Synonymy},
url = {http://dl.acm.org/citation.cfm?id=1706543.1706548},
year = 2005
}