A Novel MetaSoft Tree-Cognitive Set Model for Evaluating Criminal Litigation Efficiency under Artificial Intelligence Ecosystems

Abstract

Artificial Intelligence (AI) is transforming judicial systems by offering tools that 
enhance legal decision-making, yet measuring the efficiency of criminal litigation processes 
within these evolving digital ecosystems remains a complex task. Traditional evaluation models 
often fail to capture the hierarchical and uncertain nature of legal data, especially when AI is 
involved in tasks like evidence analysis, case prediction, or judge-assisting tools. This study 
introduces a new mathematical model named the MetaSoft Tree-Cognitive Set (MTCS), designed 
specifically to assess the efficiency of criminal litigation in AI-driven environments. MTCS 
extends existing soft set theories by integrating hierarchical attribute structures (from TreeSoft 
Set), multi-attribute interactions (from HyperSoft Set), and cognitive AI-state mapping- allowing 
for the modeling of uncertainty, legal subjectivity, and dynamic AI behavior over time. The MTCS 
model is applied in a simulated criminal case management system to evaluate litigation efficiency 
based on parameters such as case complexity, AI intervention timing, evidence ambiguity, and 
decision consistency. Through structured equations and practical demonstration, the proposed 
model not only reflects real-world legal operations but also offers policymakers a powerful tool 
for justice system optimization. The results demonstrate the MTCS’s ability to capture subtle 
changes in AI-human interaction, quantify litigation delays, and adapt to indeterminate data in 
legal environments. This research marks a step forward in blending computational intelligence 
with legal reasoning, enabling more transparent, data-informed justice practices.

DOI: 10.5281/zenodo.15399439