Consent for the use of cookies and other tools

Tools and cookies used on the website collect information about visitors in anonymized form. Your consents enable us to ensure the functioning of all website features, customize certain content specifically for you, and continuously improve the website by analyzing visits.

Types of Cookies

I consent to the website use of tools, including cookies, which ensure full functionality and an appropriate level of security. I understand that without this, the website cannot offer proper functioning, such as website navigation, customization of appearance, and access to specific parts of the website.

I consent to the website use of tools, including cookies, which collect anonymized data about website visitors. I understand that without this, website administrators cannot analyze site traffic and usage patterns to improve the user experience on the website.

Changes were successfully saved

Dr. Floriana Coppola analyzed 100 books from the general collection of the NUK, from the period 1851-2000, a transitional period in the history of stationery. She measured a total of 3,000 NIR spectra, which she analyzed with three machine learning algorithms. In this way, in cooperation with the research team, she tested how the variability of paper composition and natural decomposition affect dating models. The results showed that the common spectral properties of the cellulose and protein structure are more important for the predictions, but not the decomposition, which does not significantly affect the accuracy of the predictions.

Dr. Coppola works at the Faculty of Chemistry and Chemical Technology of the University of Ljubljana as part of the prestigious EU project of the Marie Curie program.  

The Laboratory for Heritage Science Ljubljana (Floriana Coppola, Jernej Markelj, Matija Strlič), the Faculty of Business and Economics of the University of Cagliari, Italy (Luca Frigau, Claudio Conversano) and the National and University Library of Slovenia (Jasna Malešič) participated in the research.

You can read more about the research in the article " Near-Infrared Spectroscopy and Machine Learning for Accurate Dating of Historical Books ".

Logos

  • Financira EU (en).png