Development of a visual analogy model using transfer learning techniques

Authors

  • Osowomuabe Njama-Abang
    Department of Computer Science, University of Calabar, Eta Agbor, Calabar PMB 1115, Nigeria
    https://orcid.org/0000-0001-5271-1267
  • Denis U. Ashishie
    Department of Computer Science, University of Calabar, Eta Agbor, Calabar PMB 1115, Nigeria
  • Emmanuel A. Edim
    Department of Computer Science, University of Calabar, Eta Agbor, Calabar PMB 1115, Nigeria
  • Moses A. Agana
    Department of Computer Science, University of Calabar, Eta Agbor, Calabar PMB 1115, Nigeria

Keywords:

Visual analogy models, Transfer learning, Conceptual change, Machine learning, Model development

Abstract

This study presents a novel transfer learning architecture for automated visual analogy generation to improve the effectiveness of visual analogies in facilitating learners’ comprehension of complex concepts within educational settings. While visual analogies are widely recognized as powerful tools for enhancing understanding and promoting conceptual change, existing automated models often fall short due to insufficient focus on critical design phases. To address these limitations, this research employs transfer learning technique, incorporating enriched datasets to generate visual analogies that are clear, contextually appropriate, and effective for diverse learning audiences. The computational approach integrates advanced data preprocessing, neural network architecture design, and PyTorch implementation, alongside transfer learning from Stable Diffusion. The model’s performance is assessed through a combination of quantitative metrics and qualitative evaluations. Quantitative results demonstrate that the enhanced VAM significantly outperforms the baseline, with an average FID score of 139.2091 compared to 4511.81 for the baseline, and an average CLIP similarity score of 35.3981 for the enhanced model compared to 15.6767 for the baseline. These findings underscore the potential of automated visual analogy creation to transform educational practices, offering actionable insights for educators and designers to seamlessly integrate visual analogies into teaching materials and technologies, ultimately enhancing the communication of complex ideas in classrooms.

Dimensions

[1] Y. Schur & A. Guberman, “Conceptual change of teaching among experienced teachers after studying attentive teaching”, Education Sciences 14 (2024) 231. https://doi.org/10.3390/educsci14030231.

[2] R. Hall & R. Nemirovsky, “Introduction to the special issue: Modalities of body engagement in mathematical activity and learning”, The Journal of the Learning Sciences 21 (2012) 207. http://www.jstor.org/stable/23266404.

[3] A. Sola, “Enhancing the comprehension of basic science through visual analogy”, Journal of Education and Practice 9 (2018) 8. https://api.semanticscholar.org/CorpusID:149591631.

[4] S. Glynn, “Analogies, role in science learning”, Encyclopedia of Science Education (2015) 44. https://doi.org/10.1007/978-94-007-2150-0-89.

[5] P. Gonzalez-Perez & J. J. Marrero-Galvan, “Development of a formative sequence for prospective science teachers: the challenge of improving teaching with analogies through the integration of infographics and augmented reality”, Journal of Technology and Science Education 13 (2023) 159. https://doi.org/10.3926/jotse.1919.

[6] K. Combs, H. Lu & T. J. Bihl, “Transfer learning and analogical inference: a critical comparison of algorithms, methods, and applications”, Algorithms 16 (2023) 146. https://doi.org/10.3390/a16030146.

[7] T. Webb, S. Fu, T. Bihl, K. J. Holyoak & H. Lu, “Zero-shot visual reasoning through probabilistic analogical mapping”, Nature Communications 14 (2023) 40804. https://doi.org/10.1038/s41467-023-40804-x.

[8] D. Buterez, J. P. Janet, S. J. Kiddle, D. Qglic & P. Lio, “Transfer learning with graph neural networks for improved molecular property prediction in the multi-fidelity setting”, Nature Communications 15 (2024) 1517. https://doi.org/10.1038/s41467-024-45566-8.

[9] J. Bailey, “AI in education: the leap into a new era of machine intelligence carries risks and challenges, but also plenty of promise”, Education Next 23 (2023) 28. https://www.educationnext.org/a-i-in-education-leap-into-new-era-machine-intelligence-carries-risks-challenges-promises/.

[10] M. Iman, H. R. Arabnia & K. Rasheed, “A review of deep transfer learning and recent advancements”, Technologies 11 (2023) 40. https://doi.org/10.3390/technologies11020040.

[11] A. A. Mukhlif, B. Al-Khateeb & M. A. Mohammed, “An extensive review of state-of-the-art transfer learning techniques used in medical imaging: open issues and challenges”, Journal of Intelligent Systems 31 (2022) 1085. https://doi.org/10.1515/jisys-2022-0198.

[12] K. Sezer & F. O. Karatas¸, “Research trends about analogy studies in science education: a descriptive content analysis”, Journal of Science Learning 5 (2022) 217. https://doi.org/10.17509/jsl.v5i2.39095.

[13] N. Camps, “Exploring the use of visual metaphors in teaching business and management subjects”, International Journal of Management and Applied Research 7 (2020) 340. https://doi.org/10.18646/2056.73.20-024.

[14] R. J. Amineh & H. D. Asl, “Review of constructivism and social constructivism”, Journal of Social Sciences, Literature and Languages 1 (2015) 9. https://www.academia.edu/31113252/Review-of-Constructivism-and-Social-Constructivism.

[15] D. Gentner & A. B. Markman, “Structural alignment in analogy and contrast”, Psychological Science 8 (1997) 87. https://doi.org/10.1111/j.1467-9280.1997.ep10568364.x.

[16] D. Gentner, “Structure-mapping: A theoretical framework for analogy”, Cognitive Science 7 (1983) 155. https://doi.org/10.1080/03640218308401251.

[17] D. C. Krawczyk, K. J. Holyoak & J. E. Hummel, “The one-to-one constraint in analogical mapping and inference”, Cognitive Science 29 (2008) 797. https://doi.org/10.1207/s15516709cog0000-27.

[18] S. Kornblith, M. Norouzi, H. Lee & G. Hinton, “Similarity of neural network representations revisited”, Proceedings.mlr.press 97 (2019). http://proceedings.mlr.press/v97/kornblith19a.html.

[19] M. Stieff & D. Uttal, “How much can spatial training improve STEM achievement?”, Educational Psychology Review 27 (2015) 607. https://doi.org/10.1007/s10648-015-9304-8.

[20] A. Jamrozik & D. Gentner, “Relational labeling unlocks inert knowledge”, Cognition 196 (2020) 104146. https://doi.org/10.1016/j.cognition.2019.104146.

[21] S. M. Glynn, P. J. Aubusson & S. M. Ritchie, “Learning science with analogies and models: empowering preservice elementary teachers”, Journal of Science Teacher Education 22 (2011) 51. https://doi.org/10.1007/s10972-010-9211-7.

[22] R. Li, Y. Cao, H. Tang & G. Kaiser, “Teachers’ scaffolding behavior and visual perception during cooperative learning”, International Journal of Science and Mathematics Education 22 (2023) 2. https://doi.org/10.1007/s10763-023-10379-6.

[23] K. Jaipal-Jamani & C. Angeli, “Effect of robotics on elementary preservice teachers’ self-efficacy, science learning, and computational thinking”, Journal of Science Education and Technology 26 (2017) 175. https://doi.org/10.1007/s10956-016-9663-z.

[24] J. Samani & S. C. Pan, “Interleaved practice enhances memory and problem-solving ability in undergraduate physics”, NPJ Science of Learning 6 (2021) 1. https://doi.org/10.1038/s41539-021-00110-x.

[25] I. Wayan Distrik, Z. A. Imam Supardi, B. Jatmiko & Y. Yuberti, “The effects of multiple representations-based learning in improving concept understanding and problem-solving ability”, Journal of Physics: Conference Series 1796 (2021) 012044. https://doi.org/10.1088/1742-6596/1796/1/012044.

[26] M. Bakker, J. Torbeyns, L. Verschaffel & B. De Smedt, “The mathematical, motivational, and cognitive characteristics of high mathematics achievers in primary school”, Journal of Educational Psychology 114 (2021) 5. https://doi.org/10.1037/edu0000678.

[27] R. J. Mills, P. A. Dupin-Bryant, J. D. Johnson & T. Y. Beaulieu, “Examining learning styles and perceived benefits of analogical problem construction on SQL knowledge acquisition”, Journal of Information Systems Education 26 (2015) 203. https://link.gale.com/apps/doc/A460374661/AONE?u=anon∼7cdb303b&sid=googleScholar&xid=4b655206.

[28] M. C. Wittrock, “Generative processes of comprehension”, Educational Psychologist 24 (1989) 345. https://doi.org/10.1207/s15326985ep2404-2.

[29] National Research Council, Improving adult literacy instruction: supporting learning and motivation, The National Academies Press, Washington, D.C., USA, 2012. https://nap.nationalacademies.org/catalog/13469/improving-adult-literacy-instruction-supporting-learning-and-motivation.

[30] Y. N. Kenett, R. Gold & M. Faust, “Metaphor comprehension in low and high creative individuals”, Frontiers in Psychology 9 (2018) 482. https://doi.org/10.3389/fpsyg.2018.00482.

[31] J. Nielsen, Usability Engineering, Morgan Kaufmann, 1993. https://www.elsevier.com/books/usability-engineering/nielsen/978-0-08-052029-2.

[32] S. Esser, H. Haider, C. Lustig, T. Tanaka & K. Tanaka, “Action–effect knowledge transfers to similar effect stimuli”, Psychological Research 87 (2023) 2249. https://doi.org/10.1007/s00426-023-01800-4.

[33] Institute of Medicine and National Research Council, Transforming the Workforce for Children Birth Through Age 8, National Academies Press, 2015. https://doi.org/10.17226/19401.

[34] R. Borgo, A. Abdul-Rahman, F. Mohamed, P. W. Grant, I. Reppa, L. Floridi & M. Chen, “An empirical study on using visual metaphors in visualization”, SSRN Electronic Journal (2012) 3849224. https://doi.org/10.2139/ssrn.3849224.

[35] C. Tan, F. Sun, T. Kong, W. Zhang, C. Yang & C. Liu, “A survey on deep transfer learning”, Artificial Neural Networks and Machine Learning – ICANN (2018) 270. https://doi.org/10.1007/978-3-030-01424-7_27.

[36] J. Pennington, R. Socher & C. Manning, “GloVe: global vectors for word representation”, Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP) 14 (2014) 1532. https://aclanthology.org/D14-1162.

[37] J. Devlin, M.-W. Chang, K. Lee & K. Toutanova, “BERT: pre-training of deep bidirectional transformers for language understanding”, Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics 1 (2019) 4171. https://doi.org/10.18653/v1/n19-1423.

[38] M. Palmer, D. Gildea & P. Kingsbury, “The proposition bank: an annotated corpus of semantic roles”, Computational Linguistics 31 (2005) 71. https://doi.org/10.1162/0891201053630264.

[39] R. S. Sutton & A. Barto, Reinforcement Learning: An Introduction, The MIT Press, 2018. https://doi.org/10.7551/mitpress/11841.001.0001.

[40] I. J. Goodfellow, J. Pouget-Abadie, M. Mirza, B. Xu, D. Warde-Farley, S. Ozair, A. Courville & Y. Bengio, “Generative adversarial nets”, Proceedings of the 27th International Conference on Neural Information Processing Systems 2 (2014) 2672. https://dl.acm.org/doi/10.5555/2969033.2969125.

[41] S. Ruder, Neural transfer learning for natural language processing, Ph.D. dissertation, NUI Galway, 2019. https://researchrepository.universityofgalway.ie/entities/publication/cd6308ee-6911-460b-8ca7-27d37f44fb04.

[42] D. Abrahamson & A. Bakker, “Making sense of movement in embodied design for mathematics learning”, Cognitive Research: principles and Implications 1 (2016) 1. https://doi.org/10.1186/s41235-016-0034-3.

[43] B. Zheng, B. M. McLaren, Z. Sidortsova & J. Jordan, “The impact of animations on learning abstract concepts: evidence from a large-scale online study”, Journal of Educational Psychology 114 (2022) 709. https://doi.org/10.1037/edu0000692.

[44] D. L. Morgan, Integrating qualitative and quantitative methods: a pragmatic approach, SAGE Publications, Inc. (2014). https://doi.org/10.4135/9781544304533.

[45] X. Du & P. Sun, “Generating distant analogies facilitates relational integration: intermediary role of relational mindset and cognitive load”, Frontiers in Psychology 13 (2022). https://doi.org/10.3389/fpsyg.2022.1012081.

[46] A. Sharma, K. Sharma & A. Kumar, “Real-time emotional health detection using fine-tuned transfer networks with multimodal fusion”, Neural Computing and Applications 35 (2023) 22935. https://doi.org/10.1007/s00521-022-06913-2.

[47] Textbook question answering (TQA) dataset, by Allen Institute for AI, (2023). https://allenai.org/datasets/textbook-question-answering-tqa.

[48] Y. Sabharwal, “NephroNet: A novel program for identifying renal cell carcinoma and generating synthetic training images with convolutional neural networksand diffusion models”, ArXiv (Cornell University) (2023). https://doi.org/10.48550/arxiv.2302.05830.

Visual analogy model design

Published

2025-10-28

How to Cite

Development of a visual analogy model using transfer learning techniques. (2025). African Scientific Reports, 4(3), 304. https://doi.org/10.46481/asr.2025.4.3.304

Issue

Volume 4, Issue 3, December 2025

Section

MATHEMATICAL SCIENCES SECTION
Copyright & Licensing

Copyright (c) 2025 Osowomuabe Njama-Abang, Denis U. Ashishie, Emmanuel A. Edim, Moses A. Agana

Creative Commons License

This work is licensed under a Creative Commons Attribution 4.0 International License.

How to Cite

Development of a visual analogy model using transfer learning techniques. (2025). African Scientific Reports, 4(3), 304. https://doi.org/10.46481/asr.2025.4.3.304

Similar Articles

51-59 of 59

You may also start an advanced similarity search for this article.

Most read articles by the same author(s)