A Concept is a term or phrase identified as important/meaningful in the dataset. Concepts are colored by likelihood of relatedness/correlation.
How Concepts are Created
Concepts are terms that have a high level of occurrence and meaning within the data. You can think of a Concept as an important term.
Concepts are used in Kapiche in a few key areas:
- They're the main unit of unstructured data, capturing the essence of what is being said by customers in a dataset.
- They are used to construct the Storyboard, a visual representation of what your customers are saying in the data.
- They are the most frequently mentioned terms by customers in the data, and are usually a good starting point when digging into the data with queries.
Concept colors are purely to aid intepretation and digestion of information: they do not affect analysis results in any statistical manner. Concepts are colored based on how likely they are to be related, and you can have up to 15 color groups in an Analysis (you can change the number of groups on the Storyboard page in an Analysis).