enabling researchers, students, and the general public to optimize their searches, while also enhancing and ensuring the long-term preservation of the information contained in these collections.

#AureXus #RetrospectiveConversion #WrittenHeritage #Libraries #ISBD #AI #CulturalInnovation

✅ Reduction of human errors through assisted normalization
🧠 Refocusing of teams on domain expertise

💡 𝐂𝐨𝐧𝐜𝐥𝐮𝐬𝐢𝐨𝐧

Thanks to a more efficient AI-enhanced retrospective conversion process, AureXus has facilitated access to the heritage collections of the Roger Vailland Mediatheque,

📊𝐁𝐞𝐧𝐞𝐟𝐢𝐭𝐬 𝐨𝐛𝐬𝐞𝐫𝐯𝐞𝐝 𝐭𝐡𝐫𝐨𝐮𝐠𝐡 𝐀𝐮𝐫𝐞𝐗𝐮𝐬’ 𝐢𝐧𝐭𝐞𝐠𝐫𝐚𝐭𝐢𝐨𝐧 𝐨𝐟 𝐀𝐈

⏱️ More than a 30% reduction in processing time during manuscript data entry and structuring phases
🔍 Significant improvement in the quality of content recognition (OCR / ICR)

• Automatic encoding in UNIMARC
• Integration of the records and creation of item-level holdings by the Mediatheque in Decalog, their Integrated Library System (ILS)

• Integration of artificial intelligence modules into the processing chain, with human quality control
• Launching searches and derivation attempts on target repositories (BnF, SUDOC, etc.)
• Creation of non-derivable records
• Management and processing of anomalies via AureXus’ MEDUSA platform

🔧 𝐈𝐦𝐩𝐥𝐞𝐦𝐞𝐧𝐭𝐞𝐝 𝐟𝐮𝐧𝐜𝐭𝐢𝐨𝐧𝐚𝐥𝐢𝐭𝐢𝐞𝐬

• Scanning of the original cards
• Advanced recognition of printed and handwritten content (OCR/ICR), enhanced with AI
• Automatic extraction and structuring of data
• Compliance with bibliographic standards (ISBD)

Most of the collection processed in this project is based on handwritten information, which adds an additional layer of challenges. In this context, AureXus integrates AI into its processing chain as an assistive tool, to support both efficiency and data reliability.

The collection records, currently preserved in the form of typed and handwritten cards as well as handwritten registers, will be digitized and transformed into structured metadata compliant with the ISBD (International Standard Bibliographic Description) standard, and into records in UNIMARC format.

This project will enable researchers, students, the general public, and search engines to gain a better understanding of the content in the library’s collections and will make it easier to locate the materials they are searching for.

This transformation is happening globally, and it’s reshaping how the intellectual and cultural record of our societies remains alive and meaningful in the digital age.

We look forward to sharing this journey with you.

This shift is changing how knowledge moves through the world: how it’s preserved, how it’s discovered, and how it’s shared. It makes research faster, access broader, and fragile collections safer for the future.