Comparative Evaluation of the Rule Based Approach to Representation of Adaptation Logics

Abstract

<jats:p>Due to the rapid growth of business processes digitalization, enterprise applications cover more and more business and daily life functions thus becoming more complex. Complex enterprise applications often deal with low users’ satisfaction of usability. This problem can be solved by implementing adaptation algorithms in enterprise applications, so they can be adjusted for specific context situations and specific users’ needs. Some adaptation logics representation techniques are complex and require specific knowledge and skills to manage and modify adaptation process. In this paper rule based adaptation approach is introduced where rules are used as means to manage and modify adaptation process. Rules are easy to read and understand, thereby rule based adaptation should ensure elastic, transparent and easy administrable adaptation process. The goal of this paper is to test this statement by carrying out a comparative adaptation logics representation evaluation experiment. During the experiment participants are required to complete tasks which include different forms of adaptation logics representation (code, rules and models). Experiment results are analyzed by qualitative and quantitative measures such as users’ understandability of applications behavior when adaptation case occurs and users’ satisfaction with adaptation logics representation. Experiment results are summarized and are to be used for further development of the study.</jats:p>

The Uncertain Enterprise: Achieving Adaptation Through Digital Twins and Machine Learning Extended Abstract

“…Traditionally these approaches use a collection of numerical equations to control real-time aspects of machinery; however, the architecture of the approach is appropriate for information systems providing there is an appropriate technology that can calculate the controls based on information system behaviour. Rule Based: Rules of the form if condition then adapt-action can be used to encode knowledge about adaptation actions that are required when certain system conditions arise [3,4]. Like SPLE, this approach requires the conditions and actions for adaptation to be known beforehand, which is challenging in many cases of complex systems.…”