Updated: Jul 2, 2019
The Concept Searching underlying technology consists of a number of components including the conceptClassifier and the Taxonomy Manager. These components work in concert to discover multi-word concepts within documents that are resident in both File Shares and SharePoint. Once content is uploaded to SharePoint, the classification engine will automatically apply the tags based on the multi-word concepts to the Managed Metadata site columns.
The conceptClassifier platform is an integral part of all the conceptClassifier applications, and provides an enterprise-class technology framework, comprising metadata generation, auto-classification, and taxonomy tools that enrich and extract meaning from both structured, semi-structured, and unstructured content.
At the core of the platform is compound term processing, a technology that identifies and weights multi-word concepts, based on purely statistical analysis, independent of vocabulary, language, or grammatical style. In order to leverage compound terms in the ranking algorithms, it is necessary to understand the incremental value of higher order terms with regard to their lower order component parts. Concept Searching manages this through the use of clue ‘weights’ that can be managed in the Taxonomy Manager.
One of the main features of the Taxonomy Manager is its ease of use. The Taxonomy Manager component provides Content SME’s with the ability to create and maintain taxonomies that are stored in the SharePoint Online Term Store. Subject Matter Experts that are non-technical, can manage, monitor and modify the taxonomies as terminology and end user needs mature.
In addition to the aforementioned capabilities, Concept Searching can be used to provide user specific search results based on the users viewpoint. So different term sets can be turned on or off to refine the search results to the desired topical domain.
Concept Searching can be used for auto-notification of newly saved content to interested parties. As new documents are saved into SharePoint they are immediately classified and tagged the result of which can trigger document workflows.