Data mining and knowledge discovery are crucial in pervasive computing
systems, which are distinguished by dynamic user behaviours, contextual variations,
and different management requirements. Unlike classical data mining, which uses static
databases or data warehouses, ubiquitous data mining works at the communication
level, frequently in real time and on flowing data. Streaming data mining is supported
by a variety of technologies, each with unique requirements imposed by the nature of
pervasive data, such as variability, volume, and velocity. Data transmission time,
contextual distinction, and processing accuracy are all important factors to consider.
This study investigates the capabilities of current data streaming infrastructures and
identifies issues particular to pervasive contexts. We demonstrate the limitations of
current systems in handling real-time, adaptive mining activities using hypothetical
application scenarios and instructive case studies. These examples demonstrate crucial
gaps and growing opportunities, particularly in real-time decision making and adaptive
intelligence. To summarise, pervasive data mining is a dynamic and expanding
research subject with tremendous promise for transdisciplinary applications in social
sciences, healthcare, transportation, and smart environments.
Keywords: AI ethics, Data mining, Emotion recognition, Emotional intelligence (EI), Empathy in AI.
