Examining the Efficiency of Learning-Based Algorithms in the Process of Declaring Customs

Having the declarations used in customs procedures be submitted without errors is critical. In the face of the diversity, dynamism, and complexity of the methods used in creating this declaration, human-induced declaration files are produced erroneously. These cause many problems such as loss of lab...

Full description

Saved in:
Bibliographic Details
Main Authors: Mustafa Günerkan, Ender Şahinaslan, Önder Şahınaslan
Format: Article
Language:English
Published: Istanbul University Press 2022-12-01
Series:Acta Infologica
Subjects:
Online Access:https://cdn.istanbul.edu.tr/file/JTA6CLJ8T5/98C1855791714246A5D21D3D5CE71D87
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849315398569164800
author Mustafa Günerkan
Ender Şahinaslan
Önder Şahınaslan
author_facet Mustafa Günerkan
Ender Şahinaslan
Önder Şahınaslan
author_sort Mustafa Günerkan
collection DOAJ
description Having the declarations used in customs procedures be submitted without errors is critical. In the face of the diversity, dynamism, and complexity of the methods used in creating this declaration, human-induced declaration files are produced erroneously. These cause many problems such as loss of labor, customers, and money, as well as legal problems such as contract and legal compliance. Intelligent structures supported by current information technologies are needed to solve these problems. For this purpose, being able to use learning algorithms over big data is important in the field of customs declaration creation in the logistics industry. This study evaluates the efficiency performances of learning-based algorithms regarding the customs declaration process over 4,005,343 pieces of declaration data. According to the performance measurement results, the maximum result was achieved in the Decision Tree (75.69%) and Bagging (75.70%) algorithms with respect to the Train-test split method at a test rate of 25%. Regarding the K-Fold method, which assumes K to be equal to 10, similar success rates were obtained for the Decision Tree (75.84%) and Bagging (75.83%) algorithms. These results reveal the use of machine learning algorithms to be an effective method for detecting notification errors. This can be a resource for improving customs declaration processes and developing smart control structures, as well as for new studies to be carried out in the field.
format Article
id doaj-art-e3f210ea25274e1daf25af27ab1b7e00
institution Kabale University
issn 2602-3563
language English
publishDate 2022-12-01
publisher Istanbul University Press
record_format Article
series Acta Infologica
spelling doaj-art-e3f210ea25274e1daf25af27ab1b7e002025-08-20T03:52:08ZengIstanbul University PressActa Infologica2602-35632022-12-016217518810.26650/acin.1057060123456Examining the Efficiency of Learning-Based Algorithms in the Process of Declaring CustomsMustafa Günerkan0https://orcid.org/0000-0002-4202-2801Ender Şahinaslanhttps://orcid.org/0000-0001-8519-7612Önder Şahınaslan1https://orcid.org/0000-0003-2695-5078Maltepe Üniversitesi, Istanbul, TurkiyeMaltepe Üniversitesi, Istanbul, TurkiyeHaving the declarations used in customs procedures be submitted without errors is critical. In the face of the diversity, dynamism, and complexity of the methods used in creating this declaration, human-induced declaration files are produced erroneously. These cause many problems such as loss of labor, customers, and money, as well as legal problems such as contract and legal compliance. Intelligent structures supported by current information technologies are needed to solve these problems. For this purpose, being able to use learning algorithms over big data is important in the field of customs declaration creation in the logistics industry. This study evaluates the efficiency performances of learning-based algorithms regarding the customs declaration process over 4,005,343 pieces of declaration data. According to the performance measurement results, the maximum result was achieved in the Decision Tree (75.69%) and Bagging (75.70%) algorithms with respect to the Train-test split method at a test rate of 25%. Regarding the K-Fold method, which assumes K to be equal to 10, similar success rates were obtained for the Decision Tree (75.84%) and Bagging (75.83%) algorithms. These results reveal the use of machine learning algorithms to be an effective method for detecting notification errors. This can be a resource for improving customs declaration processes and developing smart control structures, as well as for new studies to be carried out in the field.https://cdn.istanbul.edu.tr/file/JTA6CLJ8T5/98C1855791714246A5D21D3D5CE71D87customs declarationlearning algorithmslogisticsbig data
spellingShingle Mustafa Günerkan
Ender Şahinaslan
Önder Şahınaslan
Examining the Efficiency of Learning-Based Algorithms in the Process of Declaring Customs
Acta Infologica
customs declaration
learning algorithms
logistics
big data
title Examining the Efficiency of Learning-Based Algorithms in the Process of Declaring Customs
title_full Examining the Efficiency of Learning-Based Algorithms in the Process of Declaring Customs
title_fullStr Examining the Efficiency of Learning-Based Algorithms in the Process of Declaring Customs
title_full_unstemmed Examining the Efficiency of Learning-Based Algorithms in the Process of Declaring Customs
title_short Examining the Efficiency of Learning-Based Algorithms in the Process of Declaring Customs
title_sort examining the efficiency of learning based algorithms in the process of declaring customs
topic customs declaration
learning algorithms
logistics
big data
url https://cdn.istanbul.edu.tr/file/JTA6CLJ8T5/98C1855791714246A5D21D3D5CE71D87
work_keys_str_mv AT mustafagunerkan examiningtheefficiencyoflearningbasedalgorithmsintheprocessofdeclaringcustoms
AT endersahinaslan examiningtheefficiencyoflearningbasedalgorithmsintheprocessofdeclaringcustoms
AT ondersahınaslan examiningtheefficiencyoflearningbasedalgorithmsintheprocessofdeclaringcustoms