Home > Overview > Call Classification: Categories

Call Classification: Categories

Call classification takes place after call audio is analyzed and Events are identified. During call classification, Events and metadata are used to assign one or more Categories to calls. The Category classification enables SpeechMiner to find calls with specific characteristics when users request them and to perform statistical analysis on groups of calls based on their characteristics.

Categories are defined and managed in SMART using a variety of criteria:

Categories are defined globally for the entire SpeechMiner implementation and are automatically assigned to calls during processing. When changes are made to Category definitions, the entire database of calls is re-Categorized to ensure the Category assignments of all calls match the new definitions.

SpeechMiner also supports manual Categories. Manual categories are created and named in SMART but they do not have any conditions assigned to them in SMART. Calls are assigned to these Categories manually by SpeechMiner users when they review them. For example, a manual category called For Review could be defined in SMART and then used by managers to create a group of calls that require additional handling.

In order to make it easier to manage large numbers of Categories, they can be grouped hierarchically into parent Categories and sub-Categories. For example, a parent Category could be used to group a number of Categories that are relevant to a particular sales campaign. If a call is assigned to a sub-Category, it is automatically also assigned to its parent Category.

 

See also

Programs
Events: Topics and Non-Linguistic Data
Language Support
Overview of the Interface
Typical Workflows