The combination of search and aggregation together with its semantic and statistical text-mining technologies is how Silobreaker is able to serve up results in so many different ways – and across content types – enabling you to navigate and analyze without the need for keyword-based queries. The technology is built with a single purpose in mind; to provide users with meaning, context, insight and dynamic decision-support. The results are presented through graphical representations of networks, hot spot maps, trend-charts, heat-lists, and “in-focus” entity-lists – side by side with the most relevant content. Users get contextualized views and can navigate, expand or narrow their searches without having to type in new search terms.
The illustration, to the right, outlines Silobreaker’s technology platform.
The user interface consists of a large set of analytical tools, widgets and visualizations, all of which are independent from each other but highly interoperable.
Server-side, Silobreaker’s statistical and semantic algorithms perform information extraction, cluster analysis, pattern recognition, taxonomic classification and relational analysis – in addition to traditional search engine indexing.
Silobreaker has a proprietary model for importing and working with content in a uniform way regardless of its type.
The technology IP has been gradually developed since 2002 and had its first commercial application in 2005. Patents have been granted by the USPTO and EPO around Silobreaker’s relational analysis, a way to find structure in unstructured information by deducing term-to-term associations.
