Information Retrieval Facility

The Information Retrieval Facility ( IRF short ) was a research platform and served the cooperation of experts in the field of Information Retrieval ( IR). It was founded in 2006 and had its headquarters in Vienna. The IRF was the world's first e-science system, which was entirely devoted to semantic processing of text. The staff included experts, researchers and students in the areas of information retrieval and information management. The IRF has set its 2012 activities.

Objectives

The scientific objectives of Information Retrieval Facility are:

  • Modelling innovative and subject-specific information retrieval systems for global patent document collections.
  • The research and development of an adequate technical infrastructure that enables interactive experiments with formal and mathematical concepts retrieval for very large document collections.
  • The investigation of the usability of multimodal user interfaces large-scale information retrieval systems
  • Integration of users and to facilitate their needs in the process of modeling of information retrieval systems to provide an accurate assessment.
  • The possibility of different views of patent data in response to the focus to ensure.
  • Definition of standardized methods for the evaluation of information retrieval processes in the Patent Collections
  • To get the ability to text and non-text portions of a patent in a coherent way into the handle.
  • Designing, testing and evaluation of search engines that allow structured and semi - structured documents in very large patent collections to find.
  • The integration of the temporal dimension of patent documents in retrieval strategies.
  • Improving the efficiency and accuracy of patent retrieval based on ontologies and various language techniques.
  • Improved IR techniques with which the use of unstructured queries in a patent document is possible.
  • Formal (mathematical ) Identification and specification of business relevant information in the field of Intellectual Property
  • Research on efficient scaling mechanisms in the information retrieval area, taking into account the characteristics of patent data
  • Investigating and experimenting with computing architectures for very high capacity information management
  • The creation of an open e-science platform that enables the creation and implementation of IR experiments on a common research infrastructure on a uniform and easy way.
  • The discovery and exploration of novel applications and business applications that arise from information of Intellectual Property.
  • Enabling formal information retrieval, language and semantic processing in the field of applied sciences, to bring in the global industrial context.
  • Development and integration of different information access methods research on effective methods for interactive information retrieval.

Semantic Supercomputing

Current technologies for the extraction of concepts from unstructured documents associated with intensive computing power. To allow interactive experimentation with large text corpora, the IRF has a high- performance computing (HPC ) environment for performant text mining. This is equipped with the latest technologies:

  • Multi-node cluster (currently 80 core, up to 1024)
  • Full-speed interconnect technology
  • Single System Image with large Compound Memory (currently 320 GB, up to 4 TB)
  • Fully integrated configurable computing (currently 4 FPGA Core, up to 256 )

World Patent Corpus

The objective of the IRF is to create modern information retrieval technologies a platform for patent experts. It is expected that the Information Retrieval ( IR) technologies will become the center of information technology.

The totality of all patent documents is a huge text corpus dar. patents have become a crucial issue especially for global corporations and universities. The industrial user of patent data are among the most demanding and important information professionals at all. These target groups will benefit most from a technology that helps them in the exploration of large data sets.

Multimodal interaction Number cruncher
412242
de