Scholarlink

This paper presents a DSS for human trafficking risk assessment, combining machine learning with socio-legal indicators defined by Romanian Law 678/2001 and the MNIR framework. The system combines a Random Forest classifier with socio-demographic indicators to produce case-level risk profiling and explainable outputs. By incorporating MNIR-based scoring alongside machine learning predictions, the DSS ensures both predictive accuracy and alignment with established legal frameworks. Transparency is achieved through SHAP-based feature attribution, fairness through the integration of contextual factors, and traceability through the storage of predictions in a structured database. The prototype demonstrates how AI can support timely and human-centered risk management in organizational and legal contexts.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Integrating Data Science in Decision Support Systems for Risk Management: An Application in Human Trafficking Risk Assessment

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Monica-Teodora Scăunaşu

摘要

This paper presents a DSS for human trafficking risk assessment, combining machine learning with socio-legal indicators defined by Romanian Law 678/2001 and the MNIR framework. The system combines a Random Forest classifier with socio-demographic indicators to produce case-level risk profiling and explainable outputs. By incorporating MNIR-based scoring alongside machine learning predictions, the DSS ensures both predictive accuracy and alignment with established legal frameworks. Transparency is achieved through SHAP-based feature attribution, fairness through the integration of contextual factors, and traceability through the storage of predictions in a structured database. The prototype demonstrates how AI can support timely and human-centered risk management in organizational and legal contexts.