Feature Optimization for Predicting Readability of Arabic L1 and L2

Hind Saddiki; Nizar Habash; Violetta Cavalli-Sforza; Muhamed Al Khalil

Feature Optimization for Predicting Readability of Arabic L1 and L2

Hind Saddiki, Nizar Habash, Violetta Cavalli-Sforza, Muhamed Al Khalil

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

Advances in automatic readability assessment can impact the way people consume information in a number of domains. Arabic, being a low-resource and morphologically complex language, presents numerous challenges to the task of automatic readability assessment. In this paper, we present the largest and most in-depth computational readability study for Arabic to date. We study a large set of features with varying depths, from shallow words to syntactic trees, for both L1 and L2 readability tasks. Our best L1 readability accuracy result is 94.8% (75% error reduction from a commonly used baseline). The comparable results for L2 are 72.4% (45% error reduction). We also demonstrate the added value of leveraging L1 features for L2 readability prediction.

Original language	English (US)
Title of host publication	ACL 2018 - Natural Language Processing Techniques for Educational Applications, Proceedings of the 5th Workshop
Publisher	Association for Computational Linguistics (ACL)
Pages	20-29
Number of pages	10
ISBN (Electronic)	9781948087353
State	Published - 2018
Event	ACL 2018 5th Workshop on Natural Language Processing Techniques for Educational Applications, NLPTEA 2018 - Melbourne, Australia Duration: Jul 19 2018 → …

Publication series

Name	Proceedings of the Annual Meeting of the Association for Computational Linguistics
ISSN (Print)	0736-587X

Conference

Conference	ACL 2018 5th Workshop on Natural Language Processing Techniques for Educational Applications, NLPTEA 2018
Country/Territory	Australia
City	Melbourne
Period	7/19/18 → …

ASJC Scopus subject areas

Computer Science Applications
Linguistics and Language
Language and Linguistics

Cite this

Saddiki, H., Habash, N., Cavalli-Sforza, V., & Khalil, M. A. (2018). Feature Optimization for Predicting Readability of Arabic L1 and L2. In ACL 2018 - Natural Language Processing Techniques for Educational Applications, Proceedings of the 5th Workshop (pp. 20-29). (Proceedings of the Annual Meeting of the Association for Computational Linguistics). Association for Computational Linguistics (ACL).

Feature Optimization for Predicting Readability of Arabic L1 and L2. / Saddiki, Hind; Habash, Nizar; Cavalli-Sforza, Violetta et al.
ACL 2018 - Natural Language Processing Techniques for Educational Applications, Proceedings of the 5th Workshop. Association for Computational Linguistics (ACL), 2018. p. 20-29 (Proceedings of the Annual Meeting of the Association for Computational Linguistics).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Saddiki, H, Habash, N, Cavalli-Sforza, V & Khalil, MA 2018, Feature Optimization for Predicting Readability of Arabic L1 and L2. in ACL 2018 - Natural Language Processing Techniques for Educational Applications, Proceedings of the 5th Workshop. Proceedings of the Annual Meeting of the Association for Computational Linguistics, Association for Computational Linguistics (ACL), pp. 20-29, ACL 2018 5th Workshop on Natural Language Processing Techniques for Educational Applications, NLPTEA 2018, Melbourne, Australia, 7/19/18.

Saddiki H, Habash N, Cavalli-Sforza V, Khalil MA. Feature Optimization for Predicting Readability of Arabic L1 and L2. In ACL 2018 - Natural Language Processing Techniques for Educational Applications, Proceedings of the 5th Workshop. Association for Computational Linguistics (ACL). 2018. p. 20-29. (Proceedings of the Annual Meeting of the Association for Computational Linguistics).

Saddiki, Hind ; Habash, Nizar ; Cavalli-Sforza, Violetta et al. / Feature Optimization for Predicting Readability of Arabic L1 and L2. ACL 2018 - Natural Language Processing Techniques for Educational Applications, Proceedings of the 5th Workshop. Association for Computational Linguistics (ACL), 2018. pp. 20-29 (Proceedings of the Annual Meeting of the Association for Computational Linguistics).

@inproceedings{69b85bb55ce3489c8095336ae3465da1,

title = "Feature Optimization for Predicting Readability of Arabic L1 and L2",

abstract = "Advances in automatic readability assessment can impact the way people consume information in a number of domains. Arabic, being a low-resource and morphologically complex language, presents numerous challenges to the task of automatic readability assessment. In this paper, we present the largest and most in-depth computational readability study for Arabic to date. We study a large set of features with varying depths, from shallow words to syntactic trees, for both L1 and L2 readability tasks. Our best L1 readability accuracy result is 94.8% (75% error reduction from a commonly used baseline). The comparable results for L2 are 72.4% (45% error reduction). We also demonstrate the added value of leveraging L1 features for L2 readability prediction.",

author = "Hind Saddiki and Nizar Habash and Violetta Cavalli-Sforza and Khalil, {Muhamed Al}",

note = "Publisher Copyright: {\textcopyright} 2018 Association for Computational Linguistics.; ACL 2018 5th Workshop on Natural Language Processing Techniques for Educational Applications, NLPTEA 2018 ; Conference date: 19-07-2018",

year = "2018",

language = "English (US)",

series = "Proceedings of the Annual Meeting of the Association for Computational Linguistics",

publisher = "Association for Computational Linguistics (ACL)",

pages = "20--29",

booktitle = "ACL 2018 - Natural Language Processing Techniques for Educational Applications, Proceedings of the 5th Workshop",

}

TY - GEN

T1 - Feature Optimization for Predicting Readability of Arabic L1 and L2

AU - Saddiki, Hind

AU - Habash, Nizar

AU - Cavalli-Sforza, Violetta

AU - Khalil, Muhamed Al

PY - 2018

Y1 - 2018

N2 - Advances in automatic readability assessment can impact the way people consume information in a number of domains. Arabic, being a low-resource and morphologically complex language, presents numerous challenges to the task of automatic readability assessment. In this paper, we present the largest and most in-depth computational readability study for Arabic to date. We study a large set of features with varying depths, from shallow words to syntactic trees, for both L1 and L2 readability tasks. Our best L1 readability accuracy result is 94.8% (75% error reduction from a commonly used baseline). The comparable results for L2 are 72.4% (45% error reduction). We also demonstrate the added value of leveraging L1 features for L2 readability prediction.

AB - Advances in automatic readability assessment can impact the way people consume information in a number of domains. Arabic, being a low-resource and morphologically complex language, presents numerous challenges to the task of automatic readability assessment. In this paper, we present the largest and most in-depth computational readability study for Arabic to date. We study a large set of features with varying depths, from shallow words to syntactic trees, for both L1 and L2 readability tasks. Our best L1 readability accuracy result is 94.8% (75% error reduction from a commonly used baseline). The comparable results for L2 are 72.4% (45% error reduction). We also demonstrate the added value of leveraging L1 features for L2 readability prediction.

UR - http://www.scopus.com/inward/record.url?scp=85100256532&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85100256532&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:85100256532

T3 - Proceedings of the Annual Meeting of the Association for Computational Linguistics

SP - 20

EP - 29

BT - ACL 2018 - Natural Language Processing Techniques for Educational Applications, Proceedings of the 5th Workshop

PB - Association for Computational Linguistics (ACL)

T2 - ACL 2018 5th Workshop on Natural Language Processing Techniques for Educational Applications, NLPTEA 2018

Y2 - 19 July 2018

ER -

Feature Optimization for Predicting Readability of Arabic L1 and L2

Abstract

Publication series

Conference

ASJC Scopus subject areas

Other files and links

Fingerprint

Cite this