Vision and Mission
The IRF clearly aims to become a reference laboratory for global IR research. To meet this goal, the IRF fullfils the four following conditions:
Highest competence in scientific management
The IRF invites researchers from all over the world to submit information retrieval experiments of any kind. Submissions are assessed by the Scientific Board and prioritised in accordance with their technical relevance and quality, but generally the Scientific Board aims to provide expert supportive advice to improve the work rather than negative judgements.
The Board is composed of ten members from seven different countries and four continents. They are experts renowned world-wide for their contributions to diverse areas of IR science. They have contributed more than a significant share to today’s state of the art and all of them have excelled as originators, organisers or promoters in a variety of academic societies.
Providing a standardised scientific framework
As is the case too often in other research fields, IR suffers from on the one hand researchers redeveloping systems with functionality which already exists in other laboratories, or on the other the scientific work being constrained by the particular characteristics and assumptions of available packaged solutions. This is especially the case for very high performance supercomputing facilities.
The IRF provides a uniform development infrastructure that allows users to concentrate on their own research goals. Moreover, this shared basis of operation ensures basic compatibility and interoperability of tools, allowing much more incremental research. The infrastructure extends beyond open tool boxes and support for integration to access to experimental infrastructures and reference data sets. This includes GATE, Lemur/Indri, Terrier, Mathlab compatible software to exploit parallel processing, C, Java and other standard programming tools.
Providing the reference data corpus
Assessing the quality of an information retrieval algorithm requires measuring the formal adequacy of search results. However, such evaluation of results is necessarily an approximation, in addition to being highly subjective depending on the individual searcher.
Patent searches are an exception to this rule. In this case, search requests trigger very precise responses. Though not correct in “absolute” terms, they reproducibly correspond to the current state of the art in a professional context. The so-called information professionals operate at a rather high level as they are used to delivering the basis for important and expensive decisions.
