Modified Hybridized Multi-agent Oriented Approach to Analyze Work-stress Data Providing Feedback in Real Time

Abstract This paper presents a hybridized multi-agent oriented approach to develop a model capable of classifying and linguistically grading the level of work-stress. The scope of this research is to implement the model to solve a psychological problem relating to work-stress. The model uses a neural network capability as agent to classify the work-related stress data. The fuzzy logic component then transforms the crisp output from the neural network into linguistic grade. The main idea of integrating neural network and fuzzy logic techniques was to neutralize each other's weaknesses and generate a superior hybrid solution. The work-stress data was analyzed using the model and feedback provided to the users about their stress levels in real time. The result demonstrated that using this technique, the work-stress data was classified efficiently and the measured stress level was successfully described in linguistic term (Human readable). This achievement provides the user with an automated mechanism that can render a first step towards identification, prevention and making perceptible changes to the working environment. The Intelligent Multi-Agent Decision Analyser (IMADA) uses a hybridized technique that provided better solution in terms of applicability, portability and efficiency in this particular psychological domain.