In this article
The first step to getting Text Analytics working for your project is a categorization model that fits your business and your survey verbatims. Authoring offers a managed or a self-service approach to setting up your model, or a combination of the two.
An existing model is imported: We are able to import models from other vendors into Model Builder. Note that if your model contains some vendor-specific syntax, this will need to be removed before being imported into Model Builder, and this may give you some different results. Missing syntax can be addressed through Concept Miner tags or by additional work with Boolean expressions. Our Analytics team can perform this work (which would be outlined in a Statement of Work) or customers can be trained to make updates to their model in Model Builder (this requires an Analyst who understands modeling data).
You provide an outline of a categorization model: If you have an outline of a categorization model that has been used for manual categorization or coding, this can provide a good starting point. The Forsta Analytics team can provide you with guidelines for creating your model, to ensure you’re getting the actionable insights that you need from your text analytics solution. This model is then created in Model Builder using Concept Miner tags, Boolean queries or a combination of the two..
You use a generic model:Forsta has a number of generic models that have been created by the Analytics team, which can be used as a starting point. These models will need to be adjusted so that they become fit for purpose for your business and for your survey verbatim.
You do not have a model: If you have none of the above, you can still get started quickly and easily through analysis of your data by Concept Miner and then creation of your model using Concept Miner tags and Boolean queries.
Model Template
Figure 1 - Example of a template for a particular type of survey