Understand programs, topics, and phrases
The speech and text analytics feature uses programs, topics, and phrases to define a set of instructions used to perform recognition and analysis of customer data. More specifically, programs, topics, and phrases enable speech and text analytics to automatically identify and tag interactions associated with a specific business issue and provide analytics about the customer’s motivation and goal.
To work with topics and programs, administrators must have the correct permissions. For more information, see Set topic and program permissions.
A topic is made up of phrases that represent a specific intent (for example, cancellation) and each program contains one or more topics.
- Topics – Topics are collections of phrases that indicate a business level intent. For example, if you want to identify interactions in which the customer is looking to cancel a service, you could create a topic named Cancellation with a number of phrases including: “close out my account” or “I wish to cancel” and so forth. In addition, topics help to boost the recognition of specific words and phrases in voice transcription and in the interaction overview as they adapt the underlying language model to look for organization specific language in conversations. For more information, see Work with a topic.
- Programs – Programs are packages of topics that instruct speech and text analytics which business level intents to look for in recorded conversations between interaction participants. Programs are mapped to specific queues or flows and can contain topics of varying languages and dialects. This is essential as different parts of the contact center may have different business intents of interest. For more information, see Work with a program.
- Phrases – Phrases are strings of words that outline the various ways in which a topic can be expressed. Phrases can be added one by one, or in bulk using the file upload option. Any words, including words outside of the standard dictionary, can be included in a phrase and are subsequently and automatically added to the language model for recognition. For more information, see Work with a phrase.