An Effective and Practical Probabilistic Simplified Neutrosophic Approach for Accurate and Reliable Innovative Design Evaluation for Digital Media Arts

Abstract

The evaluation and classification of digital artworks remain a challenging task due to 
the inherent uncertainty, subjectivity, and imprecision in artistic design. Traditional 
computational techniques are no longer effective and fail to capture the ambiguity and 
probabilistic behavior associated with artistic evaluation, gradually becoming untrusted for 
innovative design assessment. To that end, we propose a Probabilistic Simplified Neutrosophic 
Set (PSNS)-based convolutional art classifier, that integrates neutrosophic logic to foster the 
representational power of simple Convolutional Network classifier to be able to learn and make 
a correct decision about uncertain art evaluation scenarios. With the encapsulation of PSNS, we 
can introduce probabilistic truth, indeterminacy, and falsity values to efficiently model the 
intrinsic uncertainties in digital media art evaluation. To validate our claims, we carefully choose 
a public case of digital art classification to conduct experiments to analyze the performance of 
the proposed approach. The quantitative results prove that our approach not only accentuates 
the applicability but also can improve decision-making through evaluating and refining the 
quality of elderly care services. 

DOI: 10.5281/zenodo.14962726