The Internet of Things (IoT): A Survey of Topics and Trends using Twitter Data and Topic Modeling
This study attempts to investigate the ecology and evolution of IoT by extracting hidden trends from social media data using an unsupervised topic modeling approach and provides detailed information about popular categories or topics in IoT, the associations among those topics, and their evolution since 2014.
Abstract
Extant studies have surveyed the Internet of Things (IoT) using a qualitative, literature review-based approach. These studies are similar in that the authors used predefined categories or topics such as technologies, applications, and issues, and offer a systematic review of existing studies of IoT by focusing on those individual categories. Many of these studies are technology-focused. Our study takes a digital innovation view of IoT as a complex ecosystem of technologies, industry applications, concepts, methodologies, and social institutions, which are temporally dynamic and evolve over time. Unlike the extant review/survey studies using pre-defined categories, this study attempts to investigate the ecology and evolution of IoT by extracting hidden trends from social media data using an unsupervised topic modeling approach. The proposed quantitative, topic model-based approach provides detailed information about popular categories or topics in IoT, the associations among those topics, and their evolution since 2014. By providing these quantitative findings, this study complements the qualitative approaches used in the previous studies studying the IoT landscape.