Knowledge, collaboration, and innovation took centre stage at DBpedia Day, an integral part of SEMANTiCS 2023. Held on September 20, 2023, at the HYPERION Hotel in Leipzig, Germany, this event was a unique gathering of technology professionals, industry experts, and academic researchers with a shared passion for advancing Linked Data and Semantic AI.
DBpedia members took the stage to showcase their latest tools, applications, and technical developments. Being a member of DBpedia, the Network Institute(NI) at Vrije Universiteit(VU) Amsterdam shared insights how DBpedia can be a potential source of open data for the works done in collaboration with Cultural AI Lab. As a representative of Network Institute, Sarah Shoilee (PhD Candidate from User-centric Data Science Group), gave a presentation on illustrating NI’s collaboration with Cultural AI lab and research highlights from the lab. The presentation slides are below:
Cultural AI Lab is an initiative driven by a profound commitment to the study, design, and development of socio-technological AI systems. These systems are acutely aware of the intricate and often subjective nuances of human culture. Cultural AI is an initiative driven by CWI, KNAW Humanities Cluster, KB National Library of the Netherlands, Rijksmuseum, Netherlands Institute for Sound and Vision, TNO, VU University Amsterdam, University of Amsterdam. This lab represents a harmonious fusion of AI and the humanities, showcasing an interdisciplinary approach that seeks to revolutionise not only our understanding of human culture but also the very technology that powers our digital world.
Cultural AI Lab explores a multifaceted landscape:
- Using AI for Understanding Human Culture: It recognizes the potential for AI to unravel the complexities of human culture. The lab aims to harness AI’s analytical capabilities to gain deeper insights into cultural heritage, artefacts, and practices.
- Leveraging Humanities Expertise for AI: The lab is equally dedicated to enhancing AI technology through the knowledge and expertise derived from the humanities. By collaborating with experts from these disciplines, they strive to create AI systems that respect cultural values and ethical principles.
- Dealing with Data in Context: Understanding that data takes on different meanings and connotations within specific cultural contexts, Cultural AI Lab focuses on optimising AI systems to handle input and output data within the intended application areas. This ensures that AI respects and aligns with cultural contexts and sensitivities.
- Addressing Cultural Bias: Recognizing the presence of cultural bias in data and technology, Cultural AI Lab is committed to identifying and mitigating these biases. The lab strives to create AI technology that is sensitive to cultural variations and ethical values, thus minimising bias in its applications.
In essence, Cultural AI Lab serves as a pioneering initiative that bridges the gap between AI and human culture. It represents a proactive approach to ensure that AI both understands and respects the intricate tapestry of human culture, while fostering an environment where cultural values and ethical principles are embedded in the technology that shapes our digital landscape.
Recent Highlight from Cultural AI Lab
Some of the recent publications and interesting events in the CultralAI lab are the following:
- A Knowledge Graph of Contentious Terminology: Nesterov et. al. constructed a knowledge graph representing English and Dutch contentious terminology often used in museum object descriptions. The work aims to address harmful stereotypes found in cultural heritage collections, linking contentious terms to explanations, suggestions, and alternatives.
- Multivocal Exhibition: This work was done under a 20 hrs long HackaLOD session in a Dom Tower, Utrecht Netherlands with participation of six PhD Candidates from Cultural AI lab. This hackathon is organised by Netwerk Digitaal Erfgoed. The proposed and developed solution by the Cultural AI team is a user-centric polyvocal virtual exhibition, that empowers users to explore and understand the symbolism in cultural artefacts from different perspectives, including across religions and cultures context and WON the Polyvocal Award!
- Contextual Profiling of Charged Terms in Historical Newspapers: Brate et. al examined approximately 12 million Dutch-language articles from the Dutch National Library’s newspaper corpus. The project focuses on extracting verbs, adjectives, and compound word modifiers related to nouns to identify contextual features associated with charged terms. This helps draw parallels between known-charged terms and lesser-known terms.
- Polyvocal Knowledge Modelling for Ethnographic Heritage Object Provenance: This work, by Shoilee et. al., addresses the challenge of representing the provenance of ethnographic objects. By adapting a combined model that can express the heterogeneity and polyvocality of object provenance information, the work fosters contextualization, findability, and reusability of this knowledge.
For more information on Cultural AI Lab, please visit www.cultural-ai.nl or follow us on twitter.com/cultural_ai.
Thanks to Sarah Shoilee for this news post