IBM Content Analytics

This product appeared in the market in November, 2010, and since March, 2011 became available in Russia. It integrates functions of the known Cognos Content Analytics system, means of corporate information search in diverse sources of IBM OmniFind and is constructed based on architecture of UIMA (Unstructured Information Management Architecture). The broad spectrum of types of crude data, including relational DB of different vendors, content management systems, portals, forums, blogs, message exchange systems, correspondence of e-mail, etc. is supported. The analysis of texts allows to select in them certain entities, to reveal interrelations between them, to make classification and on this basis to estimate success of marketing campaigns, to learn the emotional nature of responses to these or those initiatives of the company or state body.

Similar modules will be required also for the analysis considering features of data domain or industry specifics of the studied content. One more task which can be solved in this way is an extraction of metadata (the author, the name, duration, a key word) from multimedia files for the purpose of their involvement in analytical processing. Requirements of Content Analytics to hardware resources: as a rule, 2-4-processor servers of standard architecture from 8 GB of RAM are used. The cost of a ready system will be determined in many respects by complexity of its setup for a specific scope which is usually performed by either specialists of the customer, or the qualified partners of IBM. The product is localized for Russian at basic lexical level.