Entropy-based syntactic tree analysis for text classification: a novel approach to distinguishing between original and translated Chinese texts

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216208%3A11320%2F25%3AP9YL4LRB" target="_blank" >RIV/00216208:11320/25:P9YL4LRB - isvavai.cz</a>

Výsledek na webu

<a href="https://www.scopus.com/inward/record.uri?eid=2-s2.0-85201389489&doi=10.1093%2fllc%2ffqae030&partnerID=40&md5=903d2c494ee2da621b0b7bbc69e53087" target="_blank" >https://www.scopus.com/inward/record.uri?eid=2-s2.0-85201389489&doi=10.1093%2fllc%2ffqae030&partnerID=40&md5=903d2c494ee2da621b0b7bbc69e53087</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1093/llc/fqae030" target="_blank" >10.1093/llc/fqae030</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Entropy-based syntactic tree analysis for text classification: a novel approach to distinguishing between original and translated Chinese texts

Popis výsledku v původním jazyce

This research focuses on classifying translated and non-translated Chinese texts by analyzing syntactic rule features, using an integrated approach of machine learning and entropy analysis. The methodology employs information entropy to gauge the complexity of syntactic rules in both text types. The methodology is based on the concept of information entropy, which serves as a quantitative measure for the complexity inherent in syntactic rules as manifested from tree-based annotations. The goal of the study is to explore whether translated Chinese texts demonstrate syntactic characteristics that are significantly different from those of non-translated texts, thereby permitting a reliable classification between the two. To do this, the research calculates information entropy values for syntactic rules in two comparable corpora, one of translated and the other of non-translated Chinese texts. Then, various machine learning models are applied to these entropy metrics to identify any significant differences between the two groups. The results show significant differences in the syntactic structures. Translated texts have a higher degree of entropy, indicating more complex syntactic constructs compared to non-translated texts. These findings contribute to our understanding of the effect of translation on language syntax, with implications for text classification and translation studies. © The Author(s) 2024.

Název v anglickém jazyce

Entropy-based syntactic tree analysis for text classification: a novel approach to distinguishing between original and translated Chinese texts

Popis výsledku anglicky

This research focuses on classifying translated and non-translated Chinese texts by analyzing syntactic rule features, using an integrated approach of machine learning and entropy analysis. The methodology employs information entropy to gauge the complexity of syntactic rules in both text types. The methodology is based on the concept of information entropy, which serves as a quantitative measure for the complexity inherent in syntactic rules as manifested from tree-based annotations. The goal of the study is to explore whether translated Chinese texts demonstrate syntactic characteristics that are significantly different from those of non-translated texts, thereby permitting a reliable classification between the two. To do this, the research calculates information entropy values for syntactic rules in two comparable corpora, one of translated and the other of non-translated Chinese texts. Then, various machine learning models are applied to these entropy metrics to identify any significant differences between the two groups. The results show significant differences in the syntactic structures. Translated texts have a higher degree of entropy, indicating more complex syntactic constructs compared to non-translated texts. These findings contribute to our understanding of the effect of translation on language syntax, with implications for text classification and translation studies. © The Author(s) 2024.

Druh

J_SC - Článek v periodiku v databázi SCOPUS

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

—

Návaznosti

—

Rok uplatnění

2024

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název periodika

Digital Scholarship in the Humanities

ISSN

2055-7671

e-ISSN

—

Svazek periodika

39

Číslo periodika v rámci svazku

3

Stát vydavatele periodika

US - Spojené státy americké

Počet stran výsledku

17

Strana od-do

984-1000

Kód UT WoS článku

—

EID výsledku v databázi Scopus

2-s2.0-85201389489