Semantic Similarity of Product and Service Names in Portuguese

The problem of conceptual comparison of names plays an important role in the field of natural language processing. In this task, the goal is to choose, among a set of names, which one refers to the same concept or object as a given input name. In this paper, we propose an algorithm for comparing na...

Full description

Bibliographic Details
Main Author: Eduardo Gonçalves
Format: Article
Language:English
Published: Centro Latinoamericano de Estudios en Informática 2023-03-01
Series:CLEI Electronic Journal
Subjects:
Online Access:https://clei.org/cleiej/index.php/cleiej/article/view/557
_version_ 1797867467418107904
author Eduardo Gonçalves
author_facet Eduardo Gonçalves
author_sort Eduardo Gonçalves
collection DOAJ
description The problem of conceptual comparison of names plays an important role in the field of natural language processing. In this task, the goal is to choose, among a set of names, which one refers to the same concept or object as a given input name. In this paper, we propose an algorithm for comparing names of products and services in Portuguese that takes account of the semantic information contained in the names. The semantic similarity between two names is calculated using information from Onto.PT, the largest public lexical ontology for the Portuguese language. Experiments were conducted on a dataset composed of 5,000 pairs of names of products and services in Portuguese. Our experimental results show that the algorithm based on Onto.PT is more effective than other well-known algorithms for name comparison, producing the highest recall and precision. Moreover, results also provide interesting insights into the advantages and disadvantages of using Onto.PT for assessing the semantic similarity of names and other kinds of short texts.
first_indexed 2024-04-09T23:41:44Z
format Article
id doaj.art-935c5eb861c14456b2dfbb72c8e56b32
institution Directory Open Access Journal
issn 0717-5000
language English
last_indexed 2024-04-09T23:41:44Z
publishDate 2023-03-01
publisher Centro Latinoamericano de Estudios en Informática
record_format Article
series CLEI Electronic Journal
spelling doaj.art-935c5eb861c14456b2dfbb72c8e56b322023-03-18T14:00:31ZengCentro Latinoamericano de Estudios en InformáticaCLEI Electronic Journal0717-50002023-03-0125310.19153/cleiej.25.3.3Semantic Similarity of Product and Service Names in PortugueseEduardo Gonçalves0ENCE/IBGE The problem of conceptual comparison of names plays an important role in the field of natural language processing. In this task, the goal is to choose, among a set of names, which one refers to the same concept or object as a given input name. In this paper, we propose an algorithm for comparing names of products and services in Portuguese that takes account of the semantic information contained in the names. The semantic similarity between two names is calculated using information from Onto.PT, the largest public lexical ontology for the Portuguese language. Experiments were conducted on a dataset composed of 5,000 pairs of names of products and services in Portuguese. Our experimental results show that the algorithm based on Onto.PT is more effective than other well-known algorithms for name comparison, producing the highest recall and precision. Moreover, results also provide interesting insights into the advantages and disadvantages of using Onto.PT for assessing the semantic similarity of names and other kinds of short texts. https://clei.org/cleiej/index.php/cleiej/article/view/557Semantic SimilarityPortugueseOntologyShort TextOnto.PT
spellingShingle Eduardo Gonçalves
Semantic Similarity of Product and Service Names in Portuguese
CLEI Electronic Journal
Semantic Similarity
Portuguese
Ontology
Short Text
Onto.PT
title Semantic Similarity of Product and Service Names in Portuguese
title_full Semantic Similarity of Product and Service Names in Portuguese
title_fullStr Semantic Similarity of Product and Service Names in Portuguese
title_full_unstemmed Semantic Similarity of Product and Service Names in Portuguese
title_short Semantic Similarity of Product and Service Names in Portuguese
title_sort semantic similarity of product and service names in portuguese
topic Semantic Similarity
Portuguese
Ontology
Short Text
Onto.PT
url https://clei.org/cleiej/index.php/cleiej/article/view/557
work_keys_str_mv AT eduardogoncalves semanticsimilarityofproductandservicenamesinportuguese