Multi-Agent Teacher Assistant, A Case Study Intended for Multi-Agent Applications.
M. Fareed Arif, Hafeez Razzaq, Yijie Han · 2007
Multi-Agent Systems (MAS), a loosely coupled network of problem solvers that work together in order to resolve problems which are beyond an individual capabilities or knowledge of each problem solver, a task which other legendary paradigms say for client-server model and many distributed environments fails to do so, pose problems for scalability, reliability etc. in the development of complex distributed applications. Multi-Agent paradigm consists of a collection of autonomous agents interact in order to achieve an overall goal where data is decentralized and computation is asynchronous hence promises to overcome the above limitations like remove network latency, provide maximum concurrency, load balancing and availability enhancement. This paper presents a case study namely Multi-Agent Teacher Assistant (MATA) reflects the true power of this emerging paradigm. By implementing MATA using MAS gives us several advantages like: Dynamic service discovery, delivery, scalability, flexible structuring, dynamically extensibility, transparency, fault tolerance, code as well as data shipping and intelligent end users nevertheless very near with real world scenarios proving to be a silver bullet for the task of automation and even taking it further to the next level. Our paper thus discusses the issues of design and development of MATA, a state of art system developed using JADE, Prometheus and Agent Oriented Software Engineering 1.