Pervasive Data Science: From Data Collection to End-User Applications

Pervasive Data Science (PDS) is an emerging paradigm that combines the Internet of Things, Pervasive Computing, and Data Science to address everyday challenges. PDS differs from traditional data science in that it harnesses data from pervasive computing deployments, which affects the way data is produced and how it can be analyzed. To date, PDS has received limited attention as an independent research domain as the research field is fragmented and scattered among many different subfields. This is due to a limited understanding of the characteristics and challenges in PDS, and a lack of end-user applications that demonstrate the benefits of PDS.
This thesis paves the way for improving the adoption of PDS by offering
(i) insights into the processes that produce data, (ii) demonstrating how
pervasive computing deployments can enable wide-range of applications by
re-purposing existing sensors and capabilities of pervasive computing devices,
and (iii) highlighting the potential benefits of Pervasive Data Science by
developing end-user applications for tackling sustainable development.