ChIA: Accessing and Analysing Cultural Images with New Technologies


Abstract: 
Through the combined application of semantic technologies and artificial intelligence (AI), the project enables improved access, enrichment and analysis of a curated set of cultural food images.
Body: 

Increasingly, applications and improvements of new digital technologies and tools are being promoted in order to make digital data more accessible and analysable in established, but also in new and innovative ways.

The ChIA project, which is funded by go! Digital NEXT GENERATION, is a cooperation between the Austrian Centre for Digital Humanities and Cultural Heritage (ACDH-CH), Dublin City University, Ireland (DCU) and Europeana Local - Austria. It acts as an experimental field for semantic tools and artificial intelligence (AI) in a digital cultural context in which the experimental setting can open up new perspectives for other networked disciplines. New technologies are tested with the help of selected images from the Europeana culture portal with regard to the topic of “food” in order to enable improved access and enhanced analysis options for cultural image data.

The project investigates whether artificial intelligence (AI) can be trained to recognize images containing cultural references and representations of “food”. Another aspect is the analysis of appealing and less appealing representations for the viewer. In addition to the existing description data (metadata), which are primarily geared towards formal description criteria (e.g. artist, date, material, genre, etc.), an automation-supported data enrichment with culture-related information on the image content is to be tested using trained AI (convolutional neural networks).

ChIA is implemented within the ACDH-CH and draws on knowledge of larger existing data infrastructures such as DARIAH or E-RIHS.

Start date: 
2019
End date: 
2021
Publisher Person: 
Amelie Dorn
Accessibility: 
Open Access
Cover_image: 
Projektverantwortliche/r: 
Person name: 
Dorn, Amelie
Is contact: 
API Output Type: