A Language Framework for Measuring Semantic and Syntactic Similarity for Arabic Texts

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216208%3A11320%2F25%3A3F85CGXQ" target="_blank" >RIV/00216208:11320/25:3F85CGXQ - isvavai.cz</a>

Výsledek na webu

<a href="https://www.scopus.com/inward/record.uri?eid=2-s2.0-85188898036&doi=10.1007%2fs42979-024-02691-x&partnerID=40&md5=b78c4d0c2a44025a094611d2030a6de4" target="_blank" >https://www.scopus.com/inward/record.uri?eid=2-s2.0-85188898036&doi=10.1007%2fs42979-024-02691-x&partnerID=40&md5=b78c4d0c2a44025a094611d2030a6de4</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1007/s42979-024-02691-x" target="_blank" >10.1007/s42979-024-02691-x</a>

Jazyk výsledku

angličtina

Název v původním jazyce

A Language Framework for Measuring Semantic and Syntactic Similarity for Arabic Texts

Popis výsledku v původním jazyce

A language framework for determining the similarity of two snipped texts is proposed. The edit distance concept is employed as a frame algorithm to capture syntactic and semantic similarities. In the proposed work, syntax level distances between lemma-form words are calculated, while partial edit costs are allowed to embed semantic similarity measurements. Many knowledge resources have been used, such as words’ synonyms, negation rules, and word semantic spaces. A researchable Arabic thesaurus dictionary is built in two forms, surface form and lemma form. Semantic word spaces are generated from one of the word embedding models, which represents the words in vector spaces. The algorithm is enhanced to overcome problems with different word orders between sentences by a word permutation technique that elects the best alignment of the snipped text words to yield the best matching score. The algorithm also studied the effect of negation words on textual similarity. The proposed approach was implemented to find the similarity between Arabic language texts. Results are compared with other state-of-the-art algorithms using two benchmark datasets. The experimental results show that the proposed approach achieves a higher Pearson correlation coefficient compared to other works. © The Author(s), under exclusive licence to Springer Nature Singapore Pte Ltd 2024.

Název v anglickém jazyce

A Language Framework for Measuring Semantic and Syntactic Similarity for Arabic Texts

Popis výsledku anglicky

A language framework for determining the similarity of two snipped texts is proposed. The edit distance concept is employed as a frame algorithm to capture syntactic and semantic similarities. In the proposed work, syntax level distances between lemma-form words are calculated, while partial edit costs are allowed to embed semantic similarity measurements. Many knowledge resources have been used, such as words’ synonyms, negation rules, and word semantic spaces. A researchable Arabic thesaurus dictionary is built in two forms, surface form and lemma form. Semantic word spaces are generated from one of the word embedding models, which represents the words in vector spaces. The algorithm is enhanced to overcome problems with different word orders between sentences by a word permutation technique that elects the best alignment of the snipped text words to yield the best matching score. The algorithm also studied the effect of negation words on textual similarity. The proposed approach was implemented to find the similarity between Arabic language texts. Results are compared with other state-of-the-art algorithms using two benchmark datasets. The experimental results show that the proposed approach achieves a higher Pearson correlation coefficient compared to other works. © The Author(s), under exclusive licence to Springer Nature Singapore Pte Ltd 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

SN Computer Science

ISSN

2662-995X

e-ISSN

—

Svazek periodika

5

Číslo periodika v rámci svazku

4

Stát vydavatele periodika

US - Spojené státy americké

Počet stran výsledku

14

Strana od-do

1-14

Kód UT WoS článku

—

EID výsledku v databázi Scopus

2-s2.0-85188898036