Audio Visual Ontologies / Susan Barrett

This study path asks the learner to evaluate ontology creation in relation to automated metadata creation for audio visual digital materials, and asks learners to reflect on ways to disrupt the Anglo and Western ontologies that are often embedded in these systems. 

By Susan Barrett, Director of Library Repository Services and Technology, Arizona State University


This reflection will challenge students to evaluate traditional and non-traditional ontology creation in relation to automated metadata creation of audio visual digital materials. Automating audio visual metadata extraction is critical for the search and discovery of these rich resources. Students will explore how generating metadata that is placed into traditionally Anglo, Western cultural ontologies introduces algorithmic bias and further disenfranchises marginalized communities.

Learning Objectives

The learner will:

  • Identify methods of automatic metadata creation for audio visual content.
  • Compare the ontological structure of information based in Western thought and how it relates or differs from the descriptive structure of indigenous, marginalized or non-dominant cultural communities.
  • Discuss the challenges of automatic metadata creation for audio visual materials and normalization of that metadata in a traditional ontology, e.g., LCSH or RDA.
  • Discuss or write about the development of at least two community driven ontologies and the opportunity to correct algorithmic bias in the description of audio visual materials.  

Reflection/discussion prompt

Robust descriptive metadata is critical for discovery and for understanding entity relationships. For the purpose of this exercise, focus on the structure and organization of ontologies (complex, dynamic relations between concepts). Information science is predicated on text, and the creation of rich metadata for audio visual materials requires textual analysis to describe and discover resources.

With the advancement of automatic metadata creation or extraction for a/v materials, the process of organizing that data into textual ontologies introduces opportunities to more accurately represent the culture of creator communities and of peoples depicted in moving images. Recognition of the inherent bias in current ontologies will enable you to critically analyze and implement technological solutions that attempt to overcome algorithmic inequalities.


Self-reflection Writing or Discussion

  1. Read the literature and case studies (see below)
  2. How does Noble’s discussion of the “Problems in Classifying People” (pg. 136) influence your assessment of a traditional information science ontology like LCSH?
  3. How would you adapt the automatic collection/creation of audio visual metadata to support multiple ontologies as described in Melhuish and Candler?
  4. Referencing Windchief, et. al., how will working with creator communities and traditionally marginalized peoples affect the organizational structure, display and storage of a/v metadata? How does the description of discourse and relational accountability in Indigenous cultures influence the collection and terminology employed in an ontology influence the entity relationships within an a/v ontology?
  5. Referencing Moretense, et. al., how would you work with a creator community to gather culturally relevant information that informs the design of a culturally responsible ontology?
  6. How do levels of semantic significance, described in Christel, be codified and applied to creator community collections? Should significant semantic levels be described for all communities, why or why not?



Christel, Michael G. 2009. Automated Metadata in Multimedia Information Systems: Creation, Refinement, Use in Surrogates, and Evaluation. Synthesis Lectures on Information Concepts, Retrieval, and Services. Morgan & Claypool.

D., Jen. 2018. “How Digital Leaders Front-Load Video Metadata Tagging at Scale Part 1 and 2.” Ustudio. 2018.

Mortensen, Jonathan M., Evan P. Minty, Michael Januszyk, Timothy E. Sweeney, Alan L. Rector, Natalya F. Noy, and Mark A. Musen. 2015. “Using the Wisdom of the Crowds to Find Critical Errors in Biomedical Ontologies: A Study of SNOMED CT.” Journal of the American Medical Informatics Association 22 (3): 640–48.

Melhuish, Robin, and David Candler. 2017. “Content Is King: Automated Advanced Metadata Extraction.” Wazee Digital. March 24, 2017.

Noble, Safiya Umoja. 2018. Algorithms of Oppression: How Search Engines Reinforce Racism. New York: NYU Press.

Windchief, Sweeney, Cheryl Polacek, Michael Munson, Mary Ulrich, and Jason D. Cummins. 2017. “In Reciprocity: Responses to Critiques of Indigenous Methodologies.” Qualitative Inquiry 24 (8): 532–42.