Potential study SDGs 4 - deep learning approaches to improve problem solving with machine learning inovation - Empirical and bibliometric analysis

Aug 5, 2026

Study analyzes the potential of deep learning to improve problem solving in physics learning as part of SDGs Goal 4 implementation, combining bibliometric analysis with supporting empirical data. Bibliometrics used Scopus basic data to track scientific development over the past five years through statistical indicators. Empirical data were collected from physics/science teachers and prospective teachers in East Java, Indonesia. Findings show a year-by-year increase in deep learning–related educational articles, totaling 579 records; journals and articles were the most frequent sources. Results also highlight machine learning–based deep learning as a way to strengthen problem-solving skills and support SDGs Goal 4.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages9

Reading Time23 min

AccessFree to download

Introduction

Methods

Types of research and population

Instruments and data processing

Potential study SDGs 4 of deep learning approaches to improve problem solving with machine learning inovation: Empirical and bibliometric analysis Rahmatta Thoriq Lintangesukmanjaya1, Dwikoranto 1*, Dennisa Putri Awwalina1, Rahyu Setiani2 and Lindsay N. Bergsma3 1Department of Physics Education, Universitas Negeri Surabaya, Indonesia 2Elementary School Teacher Education, Bhinneka University PGRI, Indonesia 3Social and Behavioral Sciences, Tilburg University, Netherlands Abstract. Research with the aim of analyzing the potential of deep learning in improving problem solving in physics learning is a part of the implementation of point 4 of the SDGs was carried out using bibliometric analysis and supporting empirical data. Bibliometric analysis was carried out using Scopus basic data with an evaluation of the development of science using statistical data over the past 5 years. While empirical research data was obtained from the population of physics/science teachers and prospective teachers in East Java, Indonesia. The results obtained that the trend of physics learning research with a deep learning approach in educational articles over the past 5 years has continued to increase every year. A total of 579 data were obtained with a discussion of deep learning in problem solving to improve the quality of learning. The largest type of source in this study was journals with 435 publications and the largest type of document in this study was articles with 419 publications. The novelty of the study found that a deep learning approach could be used in physics learning through the application of machine learning to improve problem solving skills as a form of effort to implement SDGs point 4. It is recommended for future researchers to integrate deep learning in physics to improve problem solving according to the global trend of adapting innovative learning through digital media with the use of machine learning. 1 Introduction Quality education is one of the points of the Sustainable Development Goals (SDGs) is aright for citizens. Knowing that the SDGs are a series of activities established by the United Nations (UN) which aim to shape the quality of life in the world in 2030 to be better. One of the goals ofthe SDGs is quality education in SDG 4 [1,2]. Improving the quality ofeducation is one of the targets in the development of the country. In the 21st century, education is required to be able to create a generation that is able to solve problems, think critically, * Corresponding author: [dwikoranto@unesa.ac.id](dwikoranto@unesa.ac.id) © The Authors, published by EDP Sciences. This is an open access article distributed under the terms of the Creative Commons Attribution License 4.0 ([https://creativecommons.org/licenses/by/4.0/](https://creativecommons.org/licenses/by/4.0/)). logically, analytically, creatively and communicatively [3,4] . This is related to the era of technology adoption that requires students to think deeply to solve problems. Therefore, problem-solving skills are needed. Problem-solving skills are the ability to identify and analyze solutions to solve problems [5] . Based on the fact that many students' problem-solving skills are still relatively low, especially in physics learning [6,7]. The main difficulty in problem solving lies in the stages of problem identification, time planning, problem exploration, solution development, and solution evaluation [5,8] . One of the learning that requires a high level of problem-solving skills is physics learning. Apart from complex learning, less than optimal learning approaches and methods also affect the improvement of students' problem-solving skills [9] . One of the relevant learning approaches in problem solving is the deep learning approach. The deep learning approach provides students with access to think deeply, analytically and critically [10, 11]. The correlation between the deep learning approach in the national curriculum and the deep le

FAQs

What is the main purpose of this study?

To analyze the potential of deep learning in improving problem-solving skills in physics learning as part of implementing SDGs point 4.

How is bibliometric analysis conducted in the research?

Bibliometric analysis uses Scopus basic data and evaluates scientific development trends using statistical data over the past five years.

What empirical data sources are used?

Empirical data are obtained from physics/science teachers and prospective teachers in East Java, Indonesia.

What do the results indicate about deep learning research trends?

The number of physics learning research articles using deep learning approaches continues to increase every year over the past five years.

Aug 5, 2026

Upload by Clementine