Bias in the digital collections of aggregated cultural data

Inna Kizhner, Melissa Terras, Maxim Rumyantsev

Research output: Contribution to journalArticlepeer-review


The paper reviews extensive literature that demonstrates the limitations and obstacles to building representative and balanced collections of aggregated cultural data. We review several sources of bias that result from a lack of balance in digital collections: bias that is inherited from physical collections and principles of collecting; bias related to the technological processes of digitization, material quality and bias that can be traced to the work of algorithms that recommend a narrowed scope of content to users; bias connected to access and online heritage policies; reduction bias when a lack of balance narrows the focus of the collection and skews the results of the humanities research. Following the previous studies, we propose that the aggregators of digitized cultural content document their epistemological choice and present the principles that govern how digital data are selected, organized and published. In doing so, the aggregators can show the limitations and constraints of the datasets that can be used for education and research in the humanities.
The journal is a closed publication but the abstract can be published online in case you need it for the repository.
Translated title of the contributionBias in the digital collections of aggregated cultural data
Original languageRussian
Pages (from-to)169-178
JournalСибирский антропологический журнал
Issue number3 (09)
Publication statusPublished - 2020


  • Google
  • google arts and culture
  • Digital Humanities
  • Digital Cultural Heritage
  • digitisation


Dive into the research topics of 'Bias in the digital collections of aggregated cultural data'. Together they form a unique fingerprint.

Cite this