EDB Retrieval

Partner: MAGNA STEYR Fahrzeugtechnik AG & Co KG

In this project a search engine for data from the automotive industry was implemented. Data from an engeneering database can be searched as well as PDF documents (norms, project documents, ...).
A simple user interface facilitates searching for relevant documents for users which are not experts in a particular engineering domain.
Typical features of a search engine (query syntax with boolean operators, fuzzy search, wildcard search, search results with text snippets and highlighted search terms, spelling correction for search queries, ...) were implemented as well as an analysis of existing search engines was performed.

top

MISTRAL- Measurable Intelligent and Reliable Semantic Extraction and Retrieval of Multimedia Data

Partner: TU-Graz

Multimedia data has a rich and complex structure in terms of inter- and intra-document references and can be an extremely valuable source of information. However, this potential is severely limited until and unless effective methods for semantic extraction and semantic-based cross-media exploration and retrieval can be devised. Today’s leading-edge techniques in this area are working well for low-level feature extraction (e.g. colour histograms), are focussing on narrow aspects of isolated collections of multimedia data, and are dealing only with single media types. MISTRAL follows the following lines of radically new research: MISTRAL will extract a large variety of semantically relevant metadata from one media type and integrate it closely with semantic concepts derived from other media types. Eventually, the results from this cross-media semantic integration will also be fed back to the semantic extraction processes of the different media types so as to enhance the quality of the results of these processes. MISTRAL will focus on most innovative, semantic-based cross-media exploration and retrieval techniques employing concepts at different semantic levels. MISTRAL addresses the specifics of multimedia data in the global, networked context employing semantic web technologies. The MISTRAL results for semantic-based multimedia retrieval will contribute to a significant improvement of today’s human-computer interaction in multimedia retrieval and exploration applications (see http://www.mistral-project.at/).

top

NTrace- Diffusion Analysis of News Data

Partner: APA DeFacto

Press agency articles or news snippets are often cited, or copied as a whole, in publications or newspapers, with use and reuse of such material following distinct temporal patterns. Identification and tracking of such references or cites is of vital interest in many areas due to copyright and marketing issues.
The NTrace project envisions automatic analysis of newspaper articles and location of similar or identical news content. Temporal patterns in news material reuse are made accessible and navigable employing methods from the field of knowledge and information visualisation. User feedback allows refinement of results over time and adaption of the analysis environment to different domains.
A requirement definition designed in the previous year will be implemented in 2005.#

top

WebAtlas- Web based User Interface for Atlas Access (Implementation Phase)

Kooperationspartner: Brockhaus Duden Neue Medien

Common web-based geographical information systems mostly serve in tasks like route planning or locating postal adresses. Accordingly, such tools operate on very specific datasets which are often incomplete or inaccurate.
The WebAtlas project targets prototypical development of an interactive, general, geographically correct world map. Information layers like satellite views, political structure or infrastructure can be displayed. A characteristic feature is the ability to zoom out to display the whole globe at once, at any time during interaction with the map. In this mode, rendering is done in 3D and the globe is drawn with correct shading and perspective. Advanced display modes like rendering of planets (i.e. Venus and Mars) based on altitude data gathered in space missions complete the project requirements.

top

XTrend- Semiautomatische Medienresonanzanalyse

Kooperationspartner: APA DeFacto

Media resonance analysis deals with quantitative and qualitative analysis of news with regard to mentions of items of interest within selected thematic contexts. For example, the subject of the analysis could be whether a certain company is mentioned rather positively or rather negatively in the context of environment protection. Currently, such analyses are associated with high manual effort and, due to rapidly increasing volume of information, are becoming ever harder to accomplish. The goal of the project is to develop methods and tools for semi and fully automatic media resonance analysis. The main focus is to develop a modular and flexible concept which allows for a very broad definition and application of mentions, actors, groups, thematic contexts and their evaluations. Also, the knowledge which is manually fed into the system while the user is working shall be integrated into the employed statistical models. In this way the system will, to a certain degree, be capable of learning from the user, and will offer the possibility to use this knowledge in different projects and for different problems.

top