VGTrust: measuring trust for volunteered geographic information

ABSTRACT

Volunteered Geographic Information (VGI) has emerged as a large, up-to-date, and easily accessible data source. VGI can allow authoritative mapping agencies to undertake continuous improvement of their own data, adding a currency dimension previously unattainable due to high associated costs. VGI also benefits scientific and social research by facilitating quick and low-cost research data capture by the public. VGI, however, through its diversity of authorship, presents a quality assurance risk to the use of this data. This research presents a formulaic model that addresses VGI quality issues, by quantifying trust in VGI. Our ‘VGTrust’ model assesses information about a data author, and the spatial and temporal trust associated with the data they create, to produce an overall VGTrust rating metric. This metric is both easy to understand and interpret. A facilitated case study, ‘Building Our Footprints’ is presented which tests the feasibility of VGTrust model in a real-world data capture exercise run by Land Information New Zealand, New Zealand’s mapping organisation. By overcoming the trust issues in VGI, this research will allow the integration of VGI and authoritative data and potentially expand the application of VGI, thereby leveraging the power of the crowd for productive and innovative re-use.

KEYWORDS:

• Volunteered Geographic Information (VGI)
• crowdsourcing
• data quality assessment
• big data
• trust

Acknowledgments

With thanks to our colleagues on the Master of Geographic Information Science who provided a sounding board, the anonymous reviewers who helped us strengthen the paper, and Land Information New Zealand (LINZ) who funded the first author’s studies. To our families, Maeve, Toby, Mel, Finn, and Oliver.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. An infrastructure to produce, share, and utilise geospatial data and information.

2. In this paper, authors are restricted to data creators. We explore options for extending the model to account for feature modifications with multiple authors later in the paper.

3. The participation in geographic activities or generation of spatial products by everyday groups of people, often with no formal expertise in that area (Heipke 2010).

4. As internal angle was not found to have a statistically significant relationship, it was not analysed in the PCA.

Additional information

Notes on contributors

Jeremy Severinsen

Jeremy Severinsen holds a Masters degree in Geographic Information Science (MGIS), with first class honours. He currently works for New Zealand's Department of Conservation and worked previously at Land Information New Zealand (LINZ).

Mairead de Roiste

Mairead de Roiste is a Senior Lecturer in Geographic Information Science and the programme director for the Masters of Geographic Information Science (MGIS) at Victoria University of Wellington. She completed her PhD at Trinity College, Dublin and an MSc in GIS at the University of Leicester.

Femke Reitsma

Femke Reitsma is a GIS consultant and previously worked at the University of Canterbury and Edinburgh University as a GIS academic. She completed her PhD at the University of Maryland College Park.

Emir Hartato

Emir Hartato completed his Masters of Geographic Information Science (MGIS) in 2017. He currently works for Planet Labs, Inc. and is a Humanitarian OpenStreetMap Team (HOT) volunteer.