Intrafind Focuses on Managing and Classifying Data Automatically

November 21, 2012

We continue our in-depth look into software publisher Intrafind this week with another focus on the many features we have found available from the company. In addition to the proven iFinder Enterprise Search product as well as the particularly useful Tagging Service, we noted other highlights on a recent navigation of the company’s Web site.

Of specific interest was TopicFinder, an automated text classification system for topic recognition and document analysis. This product allows users to automatically gather and use information which goes beyond the typical word-based content search of documents. The main purpose of the tool is to filter and manage information. The Web site explains:

“Using the TopicFinder, for example, incoming mail from customers can be automatically forwarded to the most appropriate person responsible, or depending on their content news from news tickers can be forwarded to the editorial staff responsible for sports, politics or economic affairs.

The TopicFinder can be either specially trained regarding the information needs of the customer or it can be used with a pretrained general hierarchy of topics. The tool works fully automatically. There is no need for manual tagging. The quality of the classification is very high as the TopicFinder is based on recent linguistic and mathematical / information-theoretical methods.”

We believe this automatic tool is groundbreaking in the attempt to manage and navigate Big Data, because not all data consists simply of words. We are impressed with Intrafind’s attempt to step into this territory. The enterprise data company is located in Germany and has been making such bold moves since beginning operation in 2000. The team consists of 25 specialists who provide analysis and support. For more information, please turn to the company’s homepage.

Andrea Hayden, November 21, 2012

Sponsored by, developer of Augmentext


Comments are closed.

  • Archives

  • Recent Posts

  • Meta