Bengaluru: A project of Odeuropa consortium was awarded a grant of €2.8 million (Rs 24 crore) from the EU Horizon 2020 programme Monday for its project to document key smells through history that are described through Europe.
The project is called ‘ODEUROPA: Negotiating Olfactory and Sensory Experiences in Cultural Heritage Practice and Research’.
It’s a three-year olfactory data-mining project that aims to supplement digitisation of cultural heritage in the form of object, text and image data with olfactory data. The project will use innovative AI algorithms to recreate and archive smells.
The team consists of members from various academic backgrounds like history, science, chemistry, computational linguistics, art history, policy makers, fragrance makers, and heritage organisations.
What does Europe smell like? @IngerLeemans @WillTullett @matijastrlic @sara_hlt @caroverbeek share key scents that have defined our history https://t.co/yfSHt0xsIe #odeuropa #OlfactoryHeritage #SensoryMining pic.twitter.com/4G60mGQ7RB
— Odeuropa (@odeuropa) November 17, 2020
How does it work?
The team will first train their AI network to identify where smells would be present in text and image datasets through four centuries of European history. This includes texts in multiple languages that describe smells, and scents of scenes depicted in art.
The algorithms will also document in what context smells are mentioned, what they are associated with, and what emotions they are linked to. This data will then be stored in the ‘European Olfactory Knowledge Graph’ archive.
This data will then be placed in context by cultural history researchers to create storylines to tell the tale of how smells shape history. This will include tracing the evolution of “sensory language” over time and across linguistic cultures.
The storyline resources will be prepared in multiple formats for different types of audiences, including online data as well as toolkits that can recreate the smells.
Recreation of smells is a part of emerging olfactory museology where curators seek to provide multi-sensory engagement in museums, including the use of the sense of smell.
“We will develop new, evidence-based methodologies to quantify the impact of multisensory visitor engagement, and use this data to support the implementation of policy recommendations for the recognition, promotion, presentation and safe-guarding of our olfactory heritage,” read the project’s mission statement.
The project is divided into three parts.
The first step is using computer vision and natural language processing tools to identify and trace all references to smell and scent in image, and text collections from the 16th century to the 20th century. These will be sourced from European libraries, archives and museums.
The data set is expected to have up to 4,00,000 paintings and prints, and texts in seven languages — English, German, French, Dutch, Italian, Slovene and Latin.
The next step of the process is integrating and interpreting the results from the previous step by developing semantic data structures for documenting olfactory information and putting it in historic context with appropriate, necessary knowledge.
Storylines will also establish the main elements needed to understand what different smells meant in different historical contexts.
The last part of the project will be to closely engage with data heritage professionals to weave in the olfactory experience into the experience of heritage.
The project is expected to commence in January next year.