Automatic Metadata Tagging

The process of applying Terms or “tags” from a controlled vocabulary (often in the form of a taxonomy) to an item of content is known as classification or tagging.  These tags are stored alongside the item’s content as metadata (data about data). 

The application of these tags can be done by a human reading the item, referring to the controlled vocabulary and deciding which of the terms is appropriate.

This process can be very accurate if specialist employees (indexers) are used – people with a good understanding of the subject matter and a broad knowledge of the controlled vocabulary.  However, with large volumes of content to index, or a complex multi-subject taxonomy this process becomes resource intensive, subjective and can lack granularity.

How Semaphore delivers automatic metadata tagging

Semaphore automates the process by encapsulating linguistic evidence in an ontology and applying that to content using rule-based classification and natural language processing. This delivers automatic metadata tagging.