An Analysis of Transfer Learning Methods for Multilingual Readability Assessment

Ion Madrazo Azpiazu; Maria Soledad Pera

doi:10.1145/3386392.3397605

An Analysis of Transfer Learning Methods for Multilingual Readability Assessment

Research output: Chapter in Book/Conference proceedings/Edited volume › Conference contribution › Scientific › peer-review

5 Citations (Scopus)

Abstract

Recent advances in readability assessment have lead to the introduction of multilingual strategies that can predict the reading-level of a text regardless of its language. These strategies, however, tend to be limited to just operating in different languages rather than taking any explicit advantage of the multilingual corpora they utilize. In this manuscript, we discuss the results of the in-depth empirical analysis we conducted to assess the language transfer capabilities of four different strategies for readability assessment with increasing multilingual power. Results showcase that transfer learning is a valid option for improving the performance of readability assessment, particularly in the case of typologically-similar languages and when training corpora availability is limited.

Original language	English
Title of host publication	UMAP 2020 Adjunct - Adjunct Publication of the 28th ACM Conference on User Modeling, Adaptation and Personalization
Publisher	Association for Computing Machinery (ACM)
Pages	95-100
Number of pages	6
ISBN (Electronic)	9781450367110
DOIs	https://doi.org/10.1145/3386392.3397605
Publication status	Published - 14 Jul 2020
Externally published	Yes
Event	28th ACM International Conference on User Modeling, Adaptation, and Personalization, UMAP 2020 - Genoa, Italy Duration: 14 Jul 2020 → 17 Jul 2020

Conference

Conference	28th ACM International Conference on User Modeling, Adaptation, and Personalization, UMAP 2020
Country/Territory	Italy
City	Genoa
Period	14/07/20 → 17/07/20

Keywords

multilingual
personalization
readability assessment
text analysis

Access to Document

10.1145/3386392.3397605

Cite this

@inproceedings{6c76680e6535411b8eb5cf339cab16e8,

title = "An Analysis of Transfer Learning Methods for Multilingual Readability Assessment",

abstract = "Recent advances in readability assessment have lead to the introduction of multilingual strategies that can predict the reading-level of a text regardless of its language. These strategies, however, tend to be limited to just operating in different languages rather than taking any explicit advantage of the multilingual corpora they utilize. In this manuscript, we discuss the results of the in-depth empirical analysis we conducted to assess the language transfer capabilities of four different strategies for readability assessment with increasing multilingual power. Results showcase that transfer learning is a valid option for improving the performance of readability assessment, particularly in the case of typologically-similar languages and when training corpora availability is limited.",

keywords = "multilingual, personalization, readability assessment, text analysis",

author = "{Madrazo Azpiazu}, Ion and Pera, {Maria Soledad}",

year = "2020",

month = jul,

day = "14",

doi = "10.1145/3386392.3397605",

language = "English",

pages = "95--100",

booktitle = "UMAP 2020 Adjunct - Adjunct Publication of the 28th ACM Conference on User Modeling, Adaptation and Personalization",

publisher = "Association for Computing Machinery (ACM)",

address = "United States",

note = "28th ACM International Conference on User Modeling, Adaptation, and Personalization, UMAP 2020 ; Conference date: 14-07-2020 Through 17-07-2020",

}

Madrazo Azpiazu, I & Pera, MS 2020, An Analysis of Transfer Learning Methods for Multilingual Readability Assessment. in UMAP 2020 Adjunct - Adjunct Publication of the 28th ACM Conference on User Modeling, Adaptation and Personalization. Association for Computing Machinery (ACM), pp. 95-100, 28th ACM International Conference on User Modeling, Adaptation, and Personalization, UMAP 2020, Genoa, Italy, 14/07/20. https://doi.org/10.1145/3386392.3397605

An Analysis of Transfer Learning Methods for Multilingual Readability Assessment. / Madrazo Azpiazu, Ion; Pera, Maria Soledad.
UMAP 2020 Adjunct - Adjunct Publication of the 28th ACM Conference on User Modeling, Adaptation and Personalization. Association for Computing Machinery (ACM), 2020. p. 95-100.

Research output: Chapter in Book/Conference proceedings/Edited volume › Conference contribution › Scientific › peer-review

TY - GEN

T1 - An Analysis of Transfer Learning Methods for Multilingual Readability Assessment

AU - Madrazo Azpiazu, Ion

AU - Pera, Maria Soledad

PY - 2020/7/14

Y1 - 2020/7/14

N2 - Recent advances in readability assessment have lead to the introduction of multilingual strategies that can predict the reading-level of a text regardless of its language. These strategies, however, tend to be limited to just operating in different languages rather than taking any explicit advantage of the multilingual corpora they utilize. In this manuscript, we discuss the results of the in-depth empirical analysis we conducted to assess the language transfer capabilities of four different strategies for readability assessment with increasing multilingual power. Results showcase that transfer learning is a valid option for improving the performance of readability assessment, particularly in the case of typologically-similar languages and when training corpora availability is limited.

AB - Recent advances in readability assessment have lead to the introduction of multilingual strategies that can predict the reading-level of a text regardless of its language. These strategies, however, tend to be limited to just operating in different languages rather than taking any explicit advantage of the multilingual corpora they utilize. In this manuscript, we discuss the results of the in-depth empirical analysis we conducted to assess the language transfer capabilities of four different strategies for readability assessment with increasing multilingual power. Results showcase that transfer learning is a valid option for improving the performance of readability assessment, particularly in the case of typologically-similar languages and when training corpora availability is limited.

KW - multilingual

KW - personalization

KW - readability assessment

KW - text analysis

UR - http://www.scopus.com/inward/record.url?scp=85089262931&partnerID=8YFLogxK

U2 - 10.1145/3386392.3397605

DO - 10.1145/3386392.3397605

M3 - Conference contribution

AN - SCOPUS:85089262931

SP - 95

EP - 100

BT - UMAP 2020 Adjunct - Adjunct Publication of the 28th ACM Conference on User Modeling, Adaptation and Personalization

PB - Association for Computing Machinery (ACM)

T2 - 28th ACM International Conference on User Modeling, Adaptation, and Personalization, UMAP 2020

Y2 - 14 July 2020 through 17 July 2020

ER -

An Analysis of Transfer Learning Methods for Multilingual Readability Assessment

Abstract

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this