A Neutrosophic Triplet Partial Bipolar Metric Framework for  Quantifying AI-Generated Digital Media Content Quality

Shiming Ma; Shan Lu; Quansheng Liu

Authors

Shiming Ma College of Artificial Intelligence, Nanning University, Nanning, Guangxi, 530200, China
Shan Lu College of Artificial Intelligence, Nanning University, Nanning, Guangxi, 530200, China
Quansheng Liu College of Artificial Intelligence, Nanning University, Nanning, Guangxi, 530200, China

Keywords:

Digital media evaluation, neutrosophic triplet, bipolar metric, uncertainty modeling, quality measurement

Abstract

This paper proposes a novel mathematical framework for evaluating the quality
of digital media based on the artificial intelligence using neutrosophic triplet partial
bipolar metric spaces (NTpbMS). Traditional content evaluation models often rely on
deterministic or fuzzy systems, which fail to capture the inherent uncertainty,
contradiction, and vagueness present in human perception of quality. To address this, we
define each media artifact as a neutrosophic triplet vector that captures three aspects:
perceived quality (truth), uncertainty (indeterminacy), and distortion or degradation
(falsehood). We extend the NTpbMS structure by introducing new definitions and
mathematical equations such as Neutrosophic Quality Vector (NQV), Partial Bipolar
Distance (NPBD), and Consistency Index (NQCI). The paper presents detailed
derivations, formal proofs, and several fully calculated examples demonstrating how the
model evaluates and compares digital artifacts. The proposed model effectively quantifies
complex quality features and opens new avenues for rigorous, uncertainty-aware digital
media evaluation based on the artificial intelligence.

DOI: 10.5281/zenodo.16496921