Integrating Named Entity Recognition and Important Term Mining into the Conatix Online Business Inte

[Amin Alizadeh]

Conatix is a Berlin- and London-based startup and a spin-off project of the Humboldt-Universität zu Berlin developing an online semi-automated business intelligence (business research) system based on recent advances in machine learning.

The primary use-case for the Conatix semi-automated business intelligence system is the department of a large or medium-sized company or financial institution engaged in external research (sourcing and structuring new information that comes from outside the organization).  This could include the new product development department evaluating new product ideas, the marketing department exploring new market segments and competitors, the strategy or planning department building future macroeconomic scenarios, the legal or compliance department assessing new and existing regulatory requirements, or the investment analysis department analyzing new investment opportunities.  In all of these cases, one or more researchers will use the business intelligence system to organize their research work and to store and share their research results.

Once documents (HTML pages, PDFs, text documents or other formats available on the web and on company intranets) have been discovered by the Conatix business intelligence system, Stanbol can aid in identifying important terms, which can in turn serve as an input for discovery and relevance classification of further documents, in addition to inputs and feedback provided manually by users.

Stanbol and VIE integration into the beta prototype of the Conatix semi-automated business intelligence system will focus on named entity recognition and important term mining.

Conatix has a working alpha prototype of its system and is now completing the beta version.  The present project provides an opportunity to integrate powerful IKS functionality into our system even in the beta prototype stage.

An alpha demo of the current Conatix system (presently in stealth mode) can be shown online upon request.