EUVIMEDEUVIMED entdecken
Crossref

EXPLAINABLE BUSINESS ANALYTICS THROUGH HYBRID PSO GENETIC OPTIMIZATION AND DEEP SEMANTIC LATENT INTELLIGENCE FRAMEWORK

K. Grace Mani

ICTACT Journal on Soft Computing · 01.07.2026 · 17 (2) · 4276-4285

DOI: 10.21917/ijsc.2026.0589 ↗

Kurz gefasst

Fachartikel in der Zeitschrift ICTACT Journal on Soft Computing, online erschienen am 01.07.2026. Verfasst von K. Grace Mani.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Business analytics has become an essential component of modern enterprises because the availability of large-scale organizational data has increased substantially. The use of deep learning has enabled the extraction of complex patterns from heterogeneous business datasets. However, many deep learning models operate as black-box systems, which limits transparency and reduces stakeholder trust in analytical outcomes. Explainable Artificial Intelligence (XAI) has emerged as an effective approach that addresses the need for interpretable and reliable business decision support. The major challenge in business analytics is the development of an analytical framework that achieves high predictive performance while also providing meaningful explanations. Traditional machine learning methods often fail to capture deep semantic relationships within business data, whereas existing deep learning approaches frequently lack interpretability and optimal feature selection capabilities. This limitation affects strategic decision-making and organizational confidence in automated predictions. This study proposes the Hybrid PSO-Genetic Semantic Latent Explainable Network (HPSO-GSLEN), a novel framework that integrates Particle Swarm Optimization (PSO), Genetic Algorithm (GA), Deep Semantic Latent Analysis, and Explainable Artificial Intelligence mechanisms. The PSO module identifies promising feature subsets, whereas the GA enhances the search process through adaptive crossover and mutation operations. The optimized features are processed through a deep semantic latent analysis network that extracts hidden business knowledge representations. An explainability layer generates interpretable decision insights that support managerial understanding and validation. The proposed HPSO-GSLEN framework was evaluated using an online retail business analytics dataset. Experimental results demonstrate that the framework achieves 98.91% accuracy, 98.74% precision, 98.62% recall, 98.68% F1-score, and 99.12% AUC, outperforming conventional DNN-based Business Analytics, Semantic Latent Representation Model, and GA-FSA approaches. The framework reduces feature complexity by selecting optimal semantic attributes and improves prediction reliability through explainable feature contribution analysis. The obtained results confirm that HPSO-GSLEN provides an efficient and transparent solution for advanced business analytics applications.

Zitieren

APA

Mani, K. G. (2026). EXPLAINABLE BUSINESS ANALYTICS THROUGH HYBRID PSO GENETIC OPTIMIZATION AND DEEP SEMANTIC LATENT INTELLIGENCE FRAMEWORK. ICTACT Journal on Soft Computing, 17(2), 4276-4285. https://doi.org/10.21917/ijsc.2026.0589

Vancouver

Mani KG. EXPLAINABLE BUSINESS ANALYTICS THROUGH HYBRID PSO GENETIC OPTIMIZATION AND DEEP SEMANTIC LATENT INTELLIGENCE FRAMEWORK. ICTACT Journal on Soft Computing. 2026;17(2):4276-4285. doi:10.21917/ijsc.2026.0589

BibTeX
@article{Mani2026,
  author = {Mani, K. Grace},
  title = {EXPLAINABLE BUSINESS ANALYTICS THROUGH HYBRID PSO GENETIC OPTIMIZATION AND DEEP SEMANTIC LATENT INTELLIGENCE FRAMEWORK},
  journal = {ICTACT Journal on Soft Computing},
  year = {2026},
  volume = {17},
  number = {2},
  pages = {4276--4285},
  doi = {10.21917/ijsc.2026.0589}
}
Kostenlos starten →