Pipeline for Crowdsourced IoT Data-Modeling with AI-Supported Convergence

A central challenge of today's Internet of Things (IoT) is data-interoperability. Data representations vary heavily between scenarios, domains, and vendors. Interoperability requires common standards in data representation. The IoT changes fast but successful standardization typically takes time, and is often domain-specific. This paper presents a crowdsourced IoT data modeling process. Major features of the presented processing pipeline are a public open data model repository with model validation and a data-sensitive AI-based editor support that facilitates the modeling process and fosters model convergence. A conversion of 1200 data models from the biggest available IoT data model set confirms the applicability of the approach. A user study confirms the targeted enhanced usability and high data model convergence of the approach.