Predicting phrase-level tags using entropy inspired discriminative models

Jin Young Oh, Yo Sub Han, Jungyeul Park, Jeong Won Cha

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Citation (Scopus)

Abstract

In this paper, we describe a system which predicts phrase-level tags for eojeols in Korean using entropy inspired discriminative probabilistic models such as a conditional random fields. Instead of selecting features by the intuition of user, we use a decision tree and error analysis systematically for selecting the best feature. Once we generate all available features from the corpus, we select features by using decision tree and error analysis iteratively. Experimental results show 93.90% and 49.46% accuracy for eojeols and sentences respectively. This accuracy eventually is able to improve further syntactic analysis results. We find from the results that the better meaningful features using systematic methods is good at raising performance.

Original languageEnglish
Title of host publication2011 International Conference on Information Science and Applications, ICISA 2011
DOIs
Publication statusPublished - 2011 Jul 18
Event2011 International Conference on Information Science and Applications, ICISA 2011 - Jeju Island, Korea, Republic of
Duration: 2011 Apr 262011 Apr 29

Publication series

Name2011 International Conference on Information Science and Applications, ICISA 2011

Other

Other2011 International Conference on Information Science and Applications, ICISA 2011
CountryKorea, Republic of
CityJeju Island
Period11/4/2611/4/29

Fingerprint

Decision trees
Error analysis
Entropy
Syntactics
Statistical Models

All Science Journal Classification (ASJC) codes

  • Computer Science Applications
  • Information Systems

Cite this

Oh, J. Y., Han, Y. S., Park, J., & Cha, J. W. (2011). Predicting phrase-level tags using entropy inspired discriminative models. In 2011 International Conference on Information Science and Applications, ICISA 2011 [5772402] (2011 International Conference on Information Science and Applications, ICISA 2011). https://doi.org/10.1109/ICISA.2011.5772402
Oh, Jin Young ; Han, Yo Sub ; Park, Jungyeul ; Cha, Jeong Won. / Predicting phrase-level tags using entropy inspired discriminative models. 2011 International Conference on Information Science and Applications, ICISA 2011. 2011. (2011 International Conference on Information Science and Applications, ICISA 2011).
@inproceedings{68245d9116f44c15b48e47c4a51c6fd0,
title = "Predicting phrase-level tags using entropy inspired discriminative models",
abstract = "In this paper, we describe a system which predicts phrase-level tags for eojeols in Korean using entropy inspired discriminative probabilistic models such as a conditional random fields. Instead of selecting features by the intuition of user, we use a decision tree and error analysis systematically for selecting the best feature. Once we generate all available features from the corpus, we select features by using decision tree and error analysis iteratively. Experimental results show 93.90{\%} and 49.46{\%} accuracy for eojeols and sentences respectively. This accuracy eventually is able to improve further syntactic analysis results. We find from the results that the better meaningful features using systematic methods is good at raising performance.",
author = "Oh, {Jin Young} and Han, {Yo Sub} and Jungyeul Park and Cha, {Jeong Won}",
year = "2011",
month = "7",
day = "18",
doi = "10.1109/ICISA.2011.5772402",
language = "English",
isbn = "9781424492244",
series = "2011 International Conference on Information Science and Applications, ICISA 2011",
booktitle = "2011 International Conference on Information Science and Applications, ICISA 2011",

}

Oh, JY, Han, YS, Park, J & Cha, JW 2011, Predicting phrase-level tags using entropy inspired discriminative models. in 2011 International Conference on Information Science and Applications, ICISA 2011., 5772402, 2011 International Conference on Information Science and Applications, ICISA 2011, 2011 International Conference on Information Science and Applications, ICISA 2011, Jeju Island, Korea, Republic of, 11/4/26. https://doi.org/10.1109/ICISA.2011.5772402

Predicting phrase-level tags using entropy inspired discriminative models. / Oh, Jin Young; Han, Yo Sub; Park, Jungyeul; Cha, Jeong Won.

2011 International Conference on Information Science and Applications, ICISA 2011. 2011. 5772402 (2011 International Conference on Information Science and Applications, ICISA 2011).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Predicting phrase-level tags using entropy inspired discriminative models

AU - Oh, Jin Young

AU - Han, Yo Sub

AU - Park, Jungyeul

AU - Cha, Jeong Won

PY - 2011/7/18

Y1 - 2011/7/18

N2 - In this paper, we describe a system which predicts phrase-level tags for eojeols in Korean using entropy inspired discriminative probabilistic models such as a conditional random fields. Instead of selecting features by the intuition of user, we use a decision tree and error analysis systematically for selecting the best feature. Once we generate all available features from the corpus, we select features by using decision tree and error analysis iteratively. Experimental results show 93.90% and 49.46% accuracy for eojeols and sentences respectively. This accuracy eventually is able to improve further syntactic analysis results. We find from the results that the better meaningful features using systematic methods is good at raising performance.

AB - In this paper, we describe a system which predicts phrase-level tags for eojeols in Korean using entropy inspired discriminative probabilistic models such as a conditional random fields. Instead of selecting features by the intuition of user, we use a decision tree and error analysis systematically for selecting the best feature. Once we generate all available features from the corpus, we select features by using decision tree and error analysis iteratively. Experimental results show 93.90% and 49.46% accuracy for eojeols and sentences respectively. This accuracy eventually is able to improve further syntactic analysis results. We find from the results that the better meaningful features using systematic methods is good at raising performance.

UR - http://www.scopus.com/inward/record.url?scp=79960213792&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=79960213792&partnerID=8YFLogxK

U2 - 10.1109/ICISA.2011.5772402

DO - 10.1109/ICISA.2011.5772402

M3 - Conference contribution

AN - SCOPUS:79960213792

SN - 9781424492244

T3 - 2011 International Conference on Information Science and Applications, ICISA 2011

BT - 2011 International Conference on Information Science and Applications, ICISA 2011

ER -

Oh JY, Han YS, Park J, Cha JW. Predicting phrase-level tags using entropy inspired discriminative models. In 2011 International Conference on Information Science and Applications, ICISA 2011. 2011. 5772402. (2011 International Conference on Information Science and Applications, ICISA 2011). https://doi.org/10.1109/ICISA.2011.5772402