Automatic de-identification of Data Download Packages

Boeschoten, Laura; Voorvaart, Roos; Kaandorp, Casper S.; Goorbergh, Ruben van den; Vos, Martine De

doi:10.48550/arxiv.2105.02175

Cited by 2 publications

(2 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Digital platforms typically comply to the request for a data copy by providing the person making the request a so called data download packages (DDPs; Boeschoten et al 2022b). DDPs often are .zip-files that contain files with all collected data, such as usage history, content created or consumed by the individual, or inferences made by the platform (Boeschoten et al, 2021). After obtaining their DDPs and providing informed consent, participants can donate (an extraction of) the data from these DDPs to the researcher.…”

Section: Introductionmentioning

confidence: 99%

“…These approaches differ in their way of de-identifying (i.e., anonymizing or pseudonymizing) or pre-processing the data prior to donation. To investigate social media use and well-being among adolescents, van Driel et al (2022) collected complete Instagram DDPs which were deidentified prior to further analysis (Boeschoten et al, 2021). Kmetty and Németh (2022) used Facebook DDPs to study the validity of survey answers on music preferences.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Best practices for studies using digital data donation

Carrière,

Boeschoten,

Struminskaya

et al. 2023

Preprint

View full text Add to dashboard Cite

Data donation forms an innovative and ethical approach to collection of digital trace data. It relies on the EU legislation around personal data, which mandates data controllers to provide data subjects with a copy on personal data collected on them upon request. Participants in data donation studies can request a data controller (e.g. online platform) for a copy of their digital trace data, and actively consent to donate (part of this) data to a researcher. Setting up a data donation study involves a lot of steps and considerations, which all can threaten the study quality if executed poorly. In this paper, a workflow for setting up a data donation study is introduced. This workflow is based on error sources identified in the Total Error Framework for data donation by Boeschoten and colleagues (2022), as well as experiences in earlier data donation studies by the authors. The workflow is discussed in detail and linked to challenges and considerations for each step. All steps are illustrated by three previous studies on data donation. The paper aims to provide guidelines and a starting point for researchers wanting to conduct a data donation study.

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Best practices for studies using digital data donation

Carrière,

Boeschoten,

Struminskaya

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

Best practices for studies using digital data donation

Carrière,

Boeschoten,

Struminskaya

et al. 2024

Qual Quant

View full text Add to dashboard Cite

Digital trace data form a rich, growing source of data for social sciences and humanities. Data donation offers an innovative and ethical approach to collect these digital trace data. In data donation studies, participants request a copy of the digital trace data a data controller (e.g., large digital social media or video platforms) collected about them. The European Union’s General Data Protection Regulation obliges platforms to provide such a copy. Next, the participant can choose to share (part of) this data copy with the researcher. This way, the researcher can obtain the digital trace data of interest with active consent of the participant. Setting up a data donation study involves several steps and considerations. If executed poorly, these steps might threaten a study’s quality. In this paper, we introduce a workflow for setting up a robust data donation study. This workflow is based on error sources identified in the Total Error Framework for data donation by Boeschoten et al. (2022a) as well as on experiences in earlier data donation studies by the authors. The workflow is discussed in detail and linked to challenges and considerations for each step. We aim to provide a starting point with guidelines for researchers seeking to set up and conduct a data donation study.

show abstract

Automatic de-identification of Data Download Packages

Cited by 2 publications

References 12 publications

Best practices for studies using digital data donation

Best practices for studies using digital data donation

Best practices for studies using digital data donation

Contact Info

Product

Resources

About