Enabling Knowledge Extraction on Bike Sharing Systems Throughout Open Data.

Bike Sharing Systems (BSS) have changed urban mobility patterns. Their study as part of the overall transport system in cities is attracting growing attention in recent years. Nevertheless, some deficiencies such as the lack of convention in data serving tools and the absence of historical information difficult the analysis and improvement of realistic BSS digital platforms. Additionally, other challenges related to the Big Data nature of the analysis, have hindered an integral data analysis. This paper outlines solutions for both problems, based on a sound addressing for the Big Data Extraction-Transformation-Loading (ETL) problem of storing historical BSS data. In particular, consumption tools have been provided. They not only allow handling recorded information but also allow enhancing BSS knowledge. This way the overall system can manage other relevant information (KPIs and statistics in nature). The Big Data-inspired solution proposed in this paper solves this kind of issue, showing how it can manage more data collected during a period of about six years and from twenty-seven systems. Such data have been stored and enabled for both machine-machine communication and Human-Computer Interaction.