ERCIM News 116
January 2019
Special theme Transparency in Algorithmic Decision Making
Guest editors Andreas Rauber (TU Wien and SBA), Roberto Trasarti and Fosca Giannotti (ISTI-CNR)
- Details
- Category: Research and Innovation
- Hits: 4751
by Santiago Maraggi, Joan Baixauli and Roderick McCall (LIST)
Training for radiological events is time consuming and risky. In contrast to real sources, a prototype augmented reality system lets trainees and trainers safely learn about the necessary detection, identification and decontamination steps.
- Details
- Category: Research and Innovation
- Hits: 4479
by Zoltán Fazekas, Gábor Balázs, and Péter Gáspár (MTA SZTAKI)
In a pilot-study, an urban road environment detection function was considered for smart cars, as well as for self-driving cars. The implemented artificial neural network (ANN) based algorithms use the traffic sign (TS) and/or crossroad (CR) occurrences, along a route, as input. The TS-based and the CR-based classifiers were then merged into a compound one. The way this was accomplished serves as a simple, practical example of how to build upon modularity and how to retain some degree of it in functioning ANNs.
- Details
- Category: Research and Innovation
- Hits: 4423
by Christos Georgis, George Bruseker and Eleni Tsouloucha (ICS-FORTH)
BBTalk is an online service designed to support collaborative interdisciplinary development and extension of thesauri. At present, it serves to support the curation of the BackBone Thesaurus (BBT), a meta-thesaurus for the humanities. This service allows for the transparent, community development of the BackBone Thesaurus by enabling users to submit suggestions for changes and additions to the terminology, as well as link specialist thesauri to the meta-thesaurus terms, while enabling the thesauri curators to jointly edit, add and delete terminology. This model of cooperative editing is linked to an online discussion system that allows thesauri curators to confer with one another, exchange views and ideas and finally determine any necessary changes to the BBT.
- Details
- Category: Research and Innovation
- Hits: 8518
by Anahid N.Jalali, Alexander Schindler and Bernhard Haslhofer
Data driven prognostic systems enable us to send out an early warning of machine failure in order to reduce the cost of failures and maintenance and to improve the management of the maintenance schedule. For this purpose, robust prognostic algorithms such as deep neural networks are used whose put is often difficult to interpret and comprehend. We investigate these models with the aim of moving towards a transparent and understandable model which can be applied on critical applications such as within the manufacturing industry.
- Details
- Category: Research and Innovation
- Hits: 4330
by Gerhard Chroust (Johannes Kepler University Linz) and Georg Neubauer (Austrian Institute of Technology)
Effective and efficient communication and cooperation needs a semantically precise terminology, especially in disaster management, owing to the inherent urgency, time pressure, stress and often cultural differences of interventions. The European project Driver+ aims to measure the similarities between different countries’ terminologies surrounding disaster management. Each definition is characterised by a set of “descriptors” selected from a predefined anthology (the “bag-of-words”). The number of identical/different descriptors serves as a measure of the semantic similarity/difference of individual definitions and is translated into a numeric “degree of similarity”. The translation considers logical and intuitive aspects. Human judgment and mechanical derivation in the process are clearly separated and identified. By exchanging the ontology this method will also be applicable to other domains.
- Details
- Category: Research and Innovation
- Hits: 3503
by Ricard Munné Caldés (ATOS)
The Science2Society project improves collaboration between science and industry by leveraging big research data through sustainable business models.
- Details
- Category: Research and Innovation
- Hits: 6252
by Christian Bauckhage, Daniel Schulz and Dirk Hecker (Fraunhofer IAIS)
Deep neural networks have pushed the boundaries of artificial intelligence but their training requires vast amounts of data and high performance hardware. While truly digitised companies easily cope with these prerequisites, traditional industries still often lack the kind of data or infrastructures the current generation of end-to-end machine learning depends on. The Fraunhofer Center for Machine Learning therefore develops novel solutions which are informed by expert knowledge. These typically require less training data and are more transparent in their decision-making processes.