Graph theory based mathematical methods for improved explainable AI

Vashist, Ansh; Joshi, Aditya; Tiwari, Sanjay; Sharma, Rakesh; Jain, Tarun; Dadheech, Pankaj

doi:10.47974/JIM-2106

Open Access Research Article

Graph theory based mathematical methods for improved explainable AI

Ansh Vashistanshvst2907203@gmail.comDepartment of Computer Science Engineering Manipal University JaipurJaipur, Rajasthan, IndiaView full profile → , Aditya Joshiajoshi2002@gmail.comDepartment of Computer Science Engineering Manipal University JaipurJaipur, Rajasthan, IndiaView full profile → , Sanjay Tiwarisanjay76tiwari@gmail.comDepartment of Computer Science Engineering Arya Institute of Engineering Technology and ManagementJaipur, Rajasthan, 302028, IndiaView full profile → , Rakesh Sharmasrak.911@gmail.comDepartment of Computer Science Engineering Vivekanand Global UniversityJaipur, Rajasthan, 303012, IndiaView full profile → , ^*Tarun JainCorresponding authortarunjainjain02@gmail.com; tarun.jain@jaipur.manipal.eduDepartment of Computer Science Engineering Manipal University JaipurDepartment of Computer Science and Engineering Manipal University Jaipur Jaipur, Rajasthan, 303007, IndiaView full profile → , Pankaj Dadheechpankajdadheech777@gmail.comDepartment of Computer Science and Engineering Swami Keshvanand Institute of Technology, Management and GramothanDepartment of Computer Science & Engineering Swami Keshvanand Institute of Technology, Management & GramothanJaipur, Rajasthan, 302017, IndiaView full profile →

* Corresponding author · click or hover a name for details

pp. 627–636Vol. 28Issue 2March 2025DOI: 10.47974/JIM-2106XML

Received:: 13 Feb 2024
Published Online:: 15 Mar 2025
Article type:: Research Article
Language:: EN
Article no.:: JIM-2106
Pages:: 627–636

Abstract

This paper introduces a novel framework aimed at improving the clarity and dependability of explanations generated by AI systems specializing in Natural Language Processing (NLP). By employing a comprehensive graph-theoretical and mathematical methodology, the proposed model systematically enhances both the accuracy and reliability of AI’s explanatory functions. Central to this methodology is the Accuracy-Reliability Enhancement Framework (AREF), which utilizes graph theory to represent and analyze the AI decision-making process. This approach enables the identification and refinement of explanation pathways that maintain high fidelity and consistent reliability across various inputs. A specific objective function is developed to meticulously balance the need for high explanation fidelity and consistency against the complexity of the explanatory model. The framework leverages advanced mathematical capabilities of graph theory, including path analysis and optimization, node importance evaluation, and key subgraph identification. These features ensure that the explanations are not only transparent and easy to understand but also based on solid mathematical principles. By integrating graph-theoretical and mathematical approaches with the goals of Explainable AI (XAI), this approach significantly advances the development of AI systems that are both comprehensible and trustworthy, addressing a crucial need in contemporary XAI techniques through a clearly defined, mathematically rigorous strategy.

Keywords

Explainable AI Natural language processing Graph theory decision Interpretability in AI Fidelity-consistency optimization

Subject Classifications

Primary 65A05Secondary 26C10

References

[1] M. Li, Y. Zhan, and Y. Ge, “Fractional distance regularized level set evolution with its application to image segmentation,” IEEE Access, vol. 8, pp. 120–130 (2020).
[2] J. R. Clough, N. Byrne, I. Oksuz, V. A. Zimmer, J. A. Schnabel, and A. P. King, “A topological loss function for deep-learning based image segmentation using persistent homology,” IEEE Trans. Pattern Anal. Mach. Intell., vol. 44, no. 12, pp. 8766–8778, Dec. (2020).
[3] A. Runacher, M. J. Kazemzadeh-Parsi, D. Di Lorenzo, V. Champaney, N. Hascoet, A. Ammar, and F. Chinesta, “Describing and modeling rough composite surfaces by using topological data analysis and fractional Brownian motion,” Polymers, vol. 15, no. 6, pp. 1449 (2023).
[4] S. Lee, H. Kim, Q. X. Lieu, and J. Lee, “CNN-based image recognition for topology optimization,” Knowl.-Based Syst., vol. 198, pp. 105887 (2020).
[5] D. Yousri, M. A. Elaziz, and S. Mirjalili, “Fractional-order calculus-based flower pollination algorithm with local search for global optimization and image segmentation,” Knowl.-Based Syst., vol. 197, pp. 992–1001 (2020).
[6] C. Biswas, T. Biswas, and R. Biswas, “Relative (p, q)-φ order and relative (p, q)-φ type oriented some growth investigations of composite p-adic entire functions,” J. Interdiscip. Math., vol. 25, no. 2, pp. 233–246 (2022).
[7] E. Courtois, Topological Data Analysis of Convolutional Neural Networks Using Depthwise Separable Convolutions, Ph.D. dissertation, University of Missouri-Saint Louis (2023).
[8] M. Nishio, M. Nishio, N. Jimbo, and K. Nakane, “Homology-based image processing for automatic classification of histopathological images of lung tissue,” Cancers, vol. 13, no. 6, pp. 1192 (2021).
[9] Slimane, Ibrahim, and Zoubir Dahmani. “Normalized fractional inequalities for continuous random variables,” Journal of Interdisciplinary Mathematics, Vol. 25.2, pp. 335-349 (2022).
[10] J. R. Clough, N. Byrne, I. Oksuz, V. A. Zimmer, J. A. Schnabel, and A. P. King, “A topological loss function for deep-learning based image segmentation using persistent homology,” IEEE Trans. Pattern Anal. Mach. Intell., vol. 44, no. 12, pp. 8766–8778, Dec. (2020).

Views: 282Downloads: 6Citations: 0