Generic nodal abstraction for enhancing human-agent collaborative model with an integrated security and trust aspects

• Main Author: Mohammed, Khudair Abbas
• Format: Thesis
• Language: English
• Published: 2019
• Subjects: QA76 Computer software
• Online Access: http://umpir.ump.edu.my/id/eprint/29963/1/Generic%20nodal%20abstraction%20for%20enhancing%20human-agent%20collaborative%20model%20with%20an%20integrated%20security.pdf; http://umpir.ump.edu.my/id/eprint/29963/

Collaborative systems are viable platforms for humans that play substantial roles in task performance. Notably, collaborative systems that are modelled based on humans and software agents entail tasks and responsibilities within system processes, covers the most suitable solutions, and improve problem-solving skills of the human users. However, the issues of modelling software agents in collaborative systems present some problems caused by the diversity of tasks and the procedures. Consequently, such issues pose critical challenges to implement scheduled tasks and workflow processes. Models, which compose of humans and systems manifest various complexities. A significant challenge is how to construct purpose-built approaches for multi-agent models that collaborate with humans based on humans’ demands and positions to reduce his/her workload daily processes. In this thesis, we propose an innovative approach for modelling a human-agent collaborative system that facilitates effective collaboration and alleviate human workflow process. This approach employs humans with multiagent systems (MAS) under applicable collaborative behavior. The applicable collaborative behavior is built based on the concept of an intelligent nodal abstraction, in which each intelligent node comprises of a human, mediator agent, normal agents, and their shared functions. The functions are assigned to the node which corresponds to those that belong to a human, normal agents and their mediator agent. The nodal abstraction is generic, which could be deployed in many domain applications. The Generic Nodal Abstraction (GNA) approach is conceived with a main node and subnodes to shape a hierarchical architecture which benefits. This concept espouses the notion that agents could be deployed to assist their human counterparts in various workflow processes and handle mundane tasks. To identify generalized tasks for the human and his/her tightly-coupled mediator and normal agents in a node, we conduct a questionnaire survey on human resource departments (HR) of different organizations (municipalities and public work and healthcare administration) soliciting information pertaining to the functions performed by their employees. To ensure the safety and accuracy of the shared information performed in the GNA approach, we integrate the approach with security and trust aspects. These aspects enable safe, efficient and precise information sharing between the nodes and increase the GNA nodes confidentiality, reliability, and trustworthy. We test and simulate the GNA approach which confirm that, (i) the GNA approach facilitates the collaboration between the humans and/or agents to perform their shared actions in a convenient cooperative levels, (ii) it reduces humans’ workload and mitigate the distributed problem-solving, and (iii) consequently, its proves that the GNA with security and trust aspect enhances the viability of human agent collaborative systems when operating in various environments.