ThoughtLab: Think with us
New ways of searching and/or browsing
The amount of online services supplying geo-spatial and temporal metadata has grown rapidly in recent years. Search engines like Europeana, but also social networks like Twitter, Flickr, and YouTube, are popular providers of masses of data. The Europeana 4D interface – e4D – enables comparative visualisation of multiple queries and supports data annotated with time span data. We implemented our design in a prototype application in the context of the European project EuropeanaConnect. It is based on a client-server architecture that charges the client with the main functionality of the system. Researchers, data-journalists, and the broad public alike can use this open source framework to explore complex data – answering both time and space-related issues.
Try the demo!
For more information, or to give feedback, please contact: Ralf Stockmann
ASSETS image search demo - The ASSETS image search demo presents a technology that searches for similar images.
The demo uses 70,000 images from Europeana data providers. All image rights remain with the individual data provider. The service allows you to perform image searches based on a content-based retrieval technology using MPEG-7 as the visual descriptor to decide on the similarity of images.
Scalability has been tested with 100 million images taken from the CoPhiR dataset. It is important to stress that despite the large dataset indexed, the system is able to respond rapidly to image similarity queries.
In order to achieve this performance, the index was built using technology developed by ISTI-CNR - approximate similarity searches in metric spaces using inverted files and content-based image retrieval based on the Lucene search engine library.
See the demo: ASSETS image search demo
For more information, or to give feedback, please contact: Giuseppe Amato.
Search Engine For Europeana - A search engine that links data together to enable improved search.
Semantic-searching – a research prototype created for Europeana by the Vrije Universiteit Amsterdam (VU.nl) – uses linked data, which gives machines the ability to make associations and put search terms into context.
For example, a search for ‘Paris' intuitively leads to connected concepts such as items in the Louvre, a Paris-based institution, or paintings portraying people named Paris, such as the Greek prince who abducted Helen of Troy. From there, more links point the way to topics like the ‘myth of Paris' or the mythical Apple of Discord - a golden apple that sparked a dispute and eventually led to the Trojan War - and then on to the forbidden apple eaten by Adam and Eve.
The semantic-search engine contains data of the Rijksmuseum Amsterdam, the Musée du Louvre in Paris, and the Rijksbureau voor Kunsthistorische Documentatie (Netherlands Institute for Art History) in The Hague.
As a work-in-progress, multilingual support for semantic-searching is limited. The prototype's interface is available in English, French and Dutch.
See the prototype: Semantic Searching Prototype
For more information, or to give feedback, please contact: Jan Wielemaker.
ASSETS 3D search demo - The ASSETS 3D search demo allows users to search for 3D models that are similar to a returned search query. In this demonstrator, users can search among 300 3D models from the Google 3D Warehouse using one model as a template for finding other models similar to it. This is made possible by analysing and indexing the shape of each 3D model and using the index as a basis for search.
From a technical perspective, the similarity between the 3D models is based on a view-based, low-level feature extraction algorithm, designed and implemented by CERTH-ITI (www.iti.gr). The indexing is developed by ISTI-CNR (www.isti.cnr.it) and is an inverted file-based indexing. This indexing method is highly scalable and will support the indexing of thousands of 3D models.
The 3D Search & Retrieval task in ASSETS is still in progress. In future versions of this demo, the user will be able to upload their own 3D models and search for similar ones or draw sketches to try to find matching models.
See the demo: ASSETS 3D search demo
For more information, or to give feedback, please contact: Michalis Lazaridis.
vieu- europeana cultural heritage visual search -
vieu is a visual search engine focusing on cultural heritage content. Starting from a random set of query suggestions, the user can get a list of matching images ranked by visual similarity. These results are retrieved from a collection of images from Europeana data providers. Along with the results, the user is presented with a direct link to the image page in the Europeana portal.
A number of different visual similarity methods are available using either local or global descriptors, with an easy way of moving back and forth between them for comparison. Depending on the method, the user may also get a graphical representation of how the similarity between the images was determined.
The indexing of the database and the computing of visual similarity and ranking are based on state-of-the-art computer vision methods. The collection contains 422K Europeana images, but the methods employed can support scaling up to millions.
See the demo: vieu visual search engine
For more information, or to give feedback, please contact: Vassilis Tzouvaras.
PATHS- Personalised Access to Cultural Heritage Spaces
The European PATHSproject has created a first prototype for interactive and personalised tour guides through digital collections. The system offers suggestions about items to look at and assist in their interpretation. It demonstrates a collection discovery environment for users which combines search, narrative pathway and visualisation techniques.
- PATHS – enable users to follow pathways (or narratives) published by other users. Users can follow a path from beginning to end or they can leave the path at any time by following recommendations and links. Users who are logged in to the system can save items to their workspace and then create, edit, publish and share pathways.
- SEARCH – supports discovery of both collection items and paths through keyword search.
- EXPLORE – enables users to explore the collections using tools which are being designed to appeal to different user preferences.
Two demonstrators have been built using test data provided by Europeana and Alinari 24 ORA. The Europeana demonstrator includes around 2 million content items from Spain and the UK. The metadata which was provided has been processed and augmented with additional indexing and linkages before being integrated into the prototype application.
More technical details on Paths are available at: http://paths-project.eu/eng/Resources
For more information, or to give feedback, contact PATHS at:http://www.paths-project.eu/eng/About/Contact-Us