“AI and the future of digital preservation”

18 June 2024, 16:00 CET (UTC+2)

English version. See also Portuguese and Spanish versions of this call for proposals. NOTE: Extended submission deadline – 15th of May!

Organizers: IFLA Information Technology Section, Artificial Intelligence Special Interest Group, and IFLA Preservation and Conservation Section

Digital preservation is the safeguarding and maintenance of digital content to ensure its accessibility and usability over time. In the era of rapidly advancing technology, preservation of  digital records and artifacts becomes crucial for retaining our cultural, historical, and scientific heritage.

The concept of digitally born documents refers to information that originates in a digital format, rather than being a digital representation of a physical document. This includes emails, blogs, social media posts, and other content created and disseminated solely in the digital realm. Ebooks represent a significant shift in how we consume literature, and their preservation is vital for future generations. Preserving these documents (as well as grey literature and other special formats) requires proactive measures to address format obsolescence and data integrity.

Preserving digital arts (including performance and immersive art) involves overcoming unique challenges presented by constantly evolving technologies and formats. Artificial intelligence (AI) plays a role in the curation and restoration of digital artworks, ensuring that the essence and intent of the artist are maintained, even as the underlying technologies change.

AI is increasingly becoming a valuable tool in digital preservation initiatves. AI algorithms can aid in the automatic categorization, tagging, and metadata creation for digital content, making it easier to manage and retrieve information. Additionally, machine learning models can contribute to the identification and mitigation of digital decay or obsolescence and can act as aids to making sense, finding and summarization and synthesis of large earlier era knowledge . AI can assist in the organization and cataloging of vast digital libraries, helping users discover and access diverse literary works with greater efficiency.

While AI presents new possibilities for digital preservation and especially now mulitimedia archives, it also brings challenges such as ethical considerations, biases in algorithms, and the need for continuous adaptation to emerging technologies. Striking a balance between innovation and ethical responsibility is crucial in leveraging artificial intelligence for the long-term preservation of digital content.

For our webinar we are looking for papers descrbing practical experience encompossing both completed pragmatic successful projects but also  beta prototypes, research and development and recent test results in these areas outlined below with particular attention to new  generation AI possibilities:

  • Automated Metadata Generation: AI algorithms can analyze digital content and automatically generate more relevant subject clusters and descriptive metadata, including keywords, tags, and contextual information. This enhances the organization and searchability of digital archives, making it easier to retrieve relevant information.
  • Content Categorization and Classification: AI-powered tools can classify digital content based on its type, format, and content. This is particularly useful for large digital archives where manual categorization would be time-consuming. AI models can identify patterns and similarities, helping to organize and categorize diverse digital materials.
  • Format Migration and File ‘Normalization’: As digital formats evolve, there is a risk of obsolescence. AI can assist in the migration and transformation of digital content from one format to another, ensuring and normalizing files so they remains accessible and usable over time. This is crucial for preserving content created in outdated or proprietary formats.
  • Digital Decay Detection: AI algorithms can detect signs of digital decay or degradation in digital files. By regularly monitoring the integrity of files, AI systems can identify and address issues such as bit rot, corruption, or deterioration, preventing the loss of valuable digital assets.
  • Automation of content appraisal: The use of AI for automating appraisal processes in digital archives involves employing machine learning algorithms and other AI techniques to analyze, categorize, and make decisions about the value, relevance, and preservation priorities of digital content within archives. It could include: content (text, image, and video) analysis, duplicate detection, pattern recognition, workflow automation, user feedback integration, etc.
  • Facial and Object Recognition: In the preservation of digital photographs and visual materials, AI-powered facial recognition and object detection can help identify and tag individuals, locations, and objects within images. This enhances the contextual information associated with digital artifacts.
  • Language Processing for Textual Content: Natural Language Processing (NLP) algorithms enable the analysis of textual content, aiding in the extraction of meaningful information. This can be particularly useful for handling large volumes of textual data, such as digitized manuscripts, books, and other written documents.
  • Digital Restoration of Media: AI can be employed for the restoration of audio and visual content. Machine learning models can analyze and reconstruct damaged or degraded digital media, improving the quality and ensuring that the original artistic intent is preserved.
  • Predictive Analysis for Preservation Planning: AI can be used for predictive analysis to anticipate potential risks to digital preservation, such as hardware failures, software obsolescence, or changing file formats. This enables organizations to develop proactive preservation plans and strategies.
  • Enhanced Access and Discovery: AI-driven recommendation systems can improve user access to digital collections by suggesting relevant content based on user preferences, search history, and content similarities. This enhances the user experience and encourages exploration of digital archives.
  • Monitoring and Security: AI-based security systems can monitor digital archives for unauthorized access, potential cyber threats, or other security risks. This helps in safeguarding the integrity and confidentiality of preserved digital content.

Important dates:

  • 15 May 2024 – Updated Deadline for submission of abstract
  • 16 May 2024 – Updated Notification of selection
  • 18 June 2024 at 16:00 CET (UTC +2) – Webinar

Submission guidelines:

Proposals must be in English and include the following information:

  • Title of proposal
  • Abstract of the proposal (No more than 500 words)
  • Name/s of presenters and positions and/or titles
  • Employer or affiliated institution
  • Contact information including email address and telephone number
  • Short biographical statement of no more than 100 words
  • Note a willingness to comply with IFLA’s Authors’ Permission
  • Presentation slides submitted with author permission form. Full papers are recommended but not obligatory and will be published in IFLA’s online Repository following webinar

 

Send Proposals to:

  • Alenka Kavčič Čolić ([email protected]) and Miguel Angel Mardero Arellano ([email protected])
  • Please contact the above email addresses if there are any questions or need for more information about this webinar.

 

Please note:

  • Presentations must be in English. Each presentation should be no more than 20 minutes in length with discussion at the end of each session.
  • Abstracts and papers should be sent in Microsoft Word file, doc or rtf. formats.
  • The webinar will be recorded and published on IFLA websites.