ERCIM News No.37 - April 1999
Eidetica - A new CWI Spin-off Company
by Henk Nieland
Eidetica is the most recent of CWIs spin-off companies (seven were established in the last five years). It was created by researchers in the Interactive Information Engineering theme, combining expertise in linguistics and mathematical clustering of dependency networks for textual domains. Eidetica provides software solutions for information managers, responsible for, eg, an intranet, a trend watching service, or a library, enabling them to easily set up high quality browsing, searching and analytic services, thus increasing both their productivity and variety of competence.
Eidetica is a Twinning venture - a concept introduced by the ministry of Economic Affairs to support emerging ICT companies with excellent office facilities, business coaching and financial support in the form of seed or venture capital. Both CWI and Twinning have a share in Eidetica. The company maintains intensive contacts with the CWI research group, which focuses on automatic creation of user friendly access and control environments for large data and information collections based on content analysis.
The unique quality of Eideticas software architecture is the ability to extract quality controlled subject maps from large collections of documents. (The term eidetica refers to the ability to recollect an earlier observation as a realistic image in one?s mind.) Using these expert supervised subject maps as a basis, Eidetica?s workbenches automatically generate intranet browsers and document delivery systems that reliably and efficiently guide their users through large data collections so as to obtain all - and only - the latest relevant information.
Eideticas products range from a full text subject indexer through the metadata agent architecture to the knowledge managers workbench, each containing the document support browser technology.
In order to do more with electronic information collections than search for a few words and wrestle ones way through endless lists of not very relevant results, Eideticas software enables browsing through document collections at a conceptual level. This is achieved through the following steps:
- merge the complete, miscellaneous document collection into a uniform format;
- read all documents to extract a dictionary of subjects
- create various maps of the collection: which documents address which subjects, what authors write about what subjects, which subjects are connected to other subjects
- quality control: visualize the constructed maps and give the information manager tools to refine them
- use the subject maps to build browsing and querying interfaces that guide the user through the collection to find precisely the right information.
Eideticas software helps improve efficiency and comfort in performing the information and knowledge managers responsibilities, such as: collecting publications responding to a specific question, matching company expertise with such questions and with outside contacts, routing the continuous intake of documents and information, deciding on who should read what, and archiving and arranging collections and providing access mechanisms for clients. It also offers expanding possibilities in the analysis of large document collections, such as personnel activity matching in a large organization, and trend watching in strategic literature.
At the core of Eideticas system is a proprietary clustering method that can efficiently find groups of related information in collections about 1000 times larger than conventional methods, and advanced methods to extract subject keywords inside documents and titles.
In a nutshell, Eideticas benefits are:
- simple and intuitive interfaces
- structured screen information
- better support of existing functionality
- more efficient performance of traditional tasks
- analysis methods
- open architecture.
For more information, see: http://www.eidetica.com
Annius V. Groenink - CEO Eidetica
Tel :+31 20 888 4126