← Back to Publications List

Intuitive Interfaces for Security Analysts Interacting with AI-Blockchain Systems

Students & Supervisors

Student Authors
Mohammad Tahmedul Haque Rafid
Bachelor of Science in Computer Science & Engineering, FST
Md. Tanvir Ahmed Badhon
Bachelor of Science in Computer Science & Engineering, FST
Tajnima Akter Shifa
Bachelor of Science in Computer Science & Engineering, FST
Tasnim Sultana
Bachelor of Science in Computer Science & Engineering, FST
Yasin Arafat
Bachelor of Science in Computer Science & Engineering, FST
Supervisors
Dr. Mahfuza Khatun
Associate Professor, Faculty, FST

Abstract

Blockchain Security Operations Centers (SOCs) depend heavily on visual dashboards to monitor AI-detected smart contract bugs and unauthorized network behavior. However, most dashboards still use red or green to indicate alert severity. This makes it difficult, and possibly ineffective, for security analysts with Color Vision Deficiency (CVD) to monitor threats. This paper presents the Accessibility Mapping Engine (AME), a novel accessibility-aware and color-blind-friendly middleware architecture for AI-driven blockchain security systems. The study aims to investigate the influence of visual cues other than color on the efficiency and precision of threat detection. To this end, we developed two rapid prototypes of dashboards, one based on color and one that is accessibility-aware. Rather than just color, our design incorporated tri-modal redundant visual encodings, including geometric shapes, distinct textures, and explicit text labels to indicate severity of alerts, status of cryptographic nodes, and abnormal system behavior. A controlled, within-subjects empirical study based on A/B testing was performed with N = 35 professional security analysts under simulated deuteranopia to compare the performance of the two interfaces. Additionally, the quantitative analysis utilizing paired-samples t-tests revealed that the AME framework decreased the latency of threat identification by 41.1% (dropping from 10.93 s [SD = 8.98] to 6.43 s [SD = 4.08]), reduced the cognitive load of the analyst, and improved their overall effectiveness in combating malicious code, especially for those with CVD; in particular, target accuracy increased by 13.9% (rising from 7.00 to 7.97 successful identifications out of 8, p < 0.001). We also aimed to show that shape- and texture-based dashboards can retain the richness of security information while making the monitoring of the blockchain inclusive, effective, and accurate. These results should help create more intuitive and accessible interfaces for security analysts working with AI-blockchain systems.

Keywords

Inclusive Security Dashboards Color Vision Deficiency (CVD) AI-Assisted Threat Triage Blockchain Visual Analytics Non-Color-Dependent UX.

Publication Details

  • Type of Publication:
  • Conference Name: International Conference on Emerging Frontiers in Advanced Sciences and Technologies (EFAST) 2026
  • Date of Conference: 27/06/2026 - 27/06/2026
  • Venue: Pabna University of Science & Technology (PUST), Dhaka - Pabna Hwy, Pabna 6600, Pabna, Bangladesh
  • Organizer: Universiti Malaysia Perlis (UniMAP), 02600 Arau, Perlis, Malaysia
View on Publisher