ORD Proposal: Open Time Series

An Open Framework for Scientific Grade Time Series Datasets

Abstract

Because publications of public data providers focus on a broader audience, their datasets are often not convenient to use for academic research. As a result, data sourcing processes are implemented redundantly across research projects and institutions, often in a not entirely automated, possibly error-prone way. The Open Time Series project addresses this problem by publishing fully reproducible routines to create time series from publicly available data. Doing so yields valuable results: First, the publication of thousands of well-tested, machine-readable time series and their metadata will reduce the data sourcing burden and improve data quality within the time series analysis and economic research communities that use Swiss data. Second, the community approach we employ, the exemplary character and the inclusive approach of our open source software promise to have spillover effects to other research communities in the longer run. The network of the KOF Swiss Economic Institute leverages the impact of our project through exchange on our learnings with data providers such as the Federal Statistical Office (FSO), the Swiss National Bank (SNB), the Swiss Secretariat of Economic Affairs (SECO) and cantonal statistical offices. Finally, the inclusion of the Open Time Series framework and datasets into teaching advances Open Research Data (ORD) principles and helps the next generation of public servants to become Open Science leaders who are familiar with Swiss data landscape.

Full title	Open Time Series - An Open Framework for Scientific Grade Time Series Datasets
Keywords	Open Data, Official Statistics, Time Series Analysis, Data Literacy, Automation, Machine Readability, Community