Search Museum Next

Crowdsourcing and AI in Cultural Heritage: Can Technology Bridge the Trust Gap?

Can Artificial Intelligence (AI) and crowdsourcing truly democratise access to cultural heritage collections, or do they introduce new complexities? In today’s digital age, the accessibility of museum collections online is paramount. However, the ease with which these collections can be discovered often hinges on the quality and nature of their descriptions. Traditionally, museums have prioritised cataloguing physical attributes—such as dimensions, materials, and creators—over the narratives and meanings behind objects. This conventional approach can impede users’ ability to search for and connect with these collections on a deeper level.

At the 2023 Digital Collections Summit, Jessica BrodeFrank from the University of Illinois Chicago delved into these challenges. She illuminated how AI and crowdsourcing can synergistically enhance the accessibility and understanding of museum collections. Her presentation underscored the necessity of harmonising technological innovation with human expertise to effectively convey the rich stories encapsulated within cultural artefacts.

What volunteer descriptions add

One innovative approach discussed by BrodeFrank is crowdsourcing, which involves enlisting volunteers to enrich museum object descriptions by adding tags or keywords. A notable example is the Tag Along with Adler project, hosted on the Zooniverse platform. BrodeFrank reported that over 4,000 volunteers tagged approximately 900 images from the Adler Planetarium’s collection, generating over 200,000 individual metadata tags. She also reported improved searchability and discovery. The contribution counts describe the scale of participation; the discovery benefit is a separate outcome reported in her account.

How AI Assists in Cataloguing

In the 2023 discussion, tools such as Google Cloud Vision API and the I-Met Tagger from the Metropolitan Museum of Art illustrated the use of AI to generate descriptive tags for museum objects. However, BrodeFrank cautioned that AI systems often mirror the biases present in their training data, potentially perpetuating cultural biases and overlooking diverse perspectives. This highlights the importance of critical oversight when integrating AI into cultural heritage contexts.

The Royal Navy museum’s work on extracting search terms from curatorial descriptions offers another route to generating access metadata.

Ethical Considerations in AI and Crowdsourcing

The integration of AI and crowdsourcing in cultural heritage raises pertinent ethical questions. Crowdsourcing depends on the voluntary contributions of individuals, who typically are not compensated. Conversely, AI development often relies on the labour of underpaid workers responsible for training datasets. Museums must navigate these ethical landscapes with transparency and fairness, ensuring that their practices respect and acknowledge the contributions of all participants.


A larger set of tags is only one part of this story. BrodeFrank’s discussion also asks whose descriptions become visible and whose perspectives a catalogue leaves out. That connects the tagging work to trust: participation matters because it can change the language through which people encounter and recognise cultural material.

Whose perspectives appear in the catalogue

Trust emerged as a central theme in BrodeFrank’s presentation. In the Tag Along with Adler project, surveys revealed that individuals from underrepresented communities, particularly Black and Hispanic groups, felt that their narratives were underrepresented in museum collections. Crowdsourcing offers a pathway to address this disparity by involving these communities directly in the description and sharing of their histories, fostering a more inclusive and representative cultural record.

The Future Landscape

BrodeFrank concluded with a compelling message: AI and crowdsourcing are not merely technical tools but integral components of a museum’s mission to disseminate knowledge and tell inclusive stories. By thoughtfully combining technological advancements with human insight, museums can enhance the accessibility and relevance of their collections for diverse audiences.

For museums charting their future course, BrodeFrank’s insights serve as both a guide and a challenge: to embrace emerging technologies while steadfastly maintaining the human connection at the core of cultural heritage preservation.

–

This article is based on the presentation “Harnessing Crowdsourcing in an AI-Driven Era” by Jessica BrodeFrank at the Digital Collections Summit 2023, organised by MuseumNext.

About the author – Manuel Charr

Manuel Charr is a journalist working in the arts and cultural sectors. With a background in marketing, Manuel is drawn to arts organizations which are prepared to try inventive ways to reach new audiences.

Related Content

How De Pont Museum bridges the gap between art and community engagement

Ischa Havens, Senior Manager of Education and Public Programme at De Pont Museum for contemporary art, shares how the museum’s monthly takeover project – PUNCH....

How Do Museums Support Community Groups in Preserving Cultural Heritage?

The Irish Community Archive Network (iCAN) helps volunteer groups document and share local heritage online. Its combination of training, technical support and connections between communities...

How to Turn an Archive into a Community Movement: Crowdsourcing Lessons from National Library of Wales

Starting in autumn 2023, around 400 volunteers transcribed the names and addresses on the 390,296 signatures of the Welsh Women’s Peace Petition of 1923–24. Most...

Subscribe to the latest museum thinking

Fresh ideas from museums around the globe in your inbox each week