Rulebases, Classification and Scoring

A Rulebase comprises a series of text analysis rules that Publisher creates from concepts and their labels defined in the model. Rulebase templates govern how Publisher combines and weights labels to generate a score for each classified document. If the score is above a threshold, a metadata tag is created for that document using the concept’s preferred label. These tags can be stored as required. For example, in the document, in a content management system, a document management system or a search engine index. The scores can also be used to sort results by relevance.