Dr Vassilis Tzouvaras (National Technical University of Athens) ‘Human-Computer collaboration for data annotation and enrichment’
Presentation on digital platforms and their usage at European Cultural Heritage level: Vassilis is a member of the EC Advisory Committee for Europeana and the DCHE Expert Group since last year. He and NTUA is a main partner in the Europeana – Digital Services Infrastructure (DSI) project. They have developed the MINT system (http://mint.image.ece.ntua.gr/redmine/projects/mint/wiki) that is being used by Europeana and a large number of Europeana projects, for metadata interoperability and the WITH system for Culture and Creativity (www.with.image.ntua.gr) which performs intelligent cultural content discovery and aggregation, semantic content enrichment and provides users with the ability to reuse content, generate personal collections, exhibitions and stories. He has been currently co-leading two big EC projects: – The CultureLabs Project, in the framework of the CULT-COOP-06-2017-Participatory approaches and social innovation in culture Programme: The project generates an ICT-based infrastructure targeting a participatory approach to social innovation, also involving immigrants’ communities. – The CrowdHeritage: Crowdsourcing Platform for Enriching Europeana Metadata Project, in the framework of CEF-TC-2017-3: The project will use ICT to evaluate the impact of crowdsourcing for Cultural Heritage quality improvement and create crowdsourcing campaigns on different topics, including European cities and landscapes and involving different target groups, specifically pupils and teachers, their families, researchers and the broader public.
Prof Stefanos Kollias (University of Lincoln), Andreas-George Stafylopatis & Ilianna Kollia‘Artificial Intelligence and Machine Learning for Cultural Heritage Content Enrichment and Search on the Web’
Cultural Heritage Institutions, such as museums, archives, libraries and galleries have been contributing to the European and UK Cultural and Creative Sectors, by digitising their content, annotating it using different metadata standards, and publishing it on the Web, through their web sites, or through domain and thematic aggregators. During the last decade many cultural institutions have participated to the creation of Europeana, i.e., the European Digital Library, contributing with more than three millions cultural objects, with text and image descriptions, that are searchable through the Europeana portal. Metadata interoperability has been a key issue in achieving searchability among these content sources and Europeana. The main target of this cultural content aggregation and sharing is boosting of innovation, by improving the way of accessing and reusing knowledge and information by large audiences of the Cultural and Creative Sectors, including young persons, researchers, young professionals and small companies. In the above framework, the target addressed in this talk/paper/panel is the use of recent technological breakthrough in Artificial Intelligence and Machine Learning for enrichment of the content of Cultural Heritage Institutions through metadata analysis, so that it can be easily searched and reused by the above audiences and the creative industries. This includes presenting and discussing current practices on metadata mappings through the use of controlled vocabularies and semantic technologies, as well as on the generation of data-driven content/metadata enrichment through the use of novel deep learning approaches. A variety of activities and projects have been or are being implemented on developing and using such technologies in different Cultural heritage fields, such as museums, libraries, archives and thematic areas such as film and TV archiving, photography and fashion. A presentation and discussion on the prospects of these activities will be tackled in the proposed framework.