A typical semantic engine analyses text based on keywords. Staffino's implemented engine doesn't require keywords for its analysis.
Our Algorithm will automatically define new topics and can compare them to drivers designated by you. Staffino's semantic engine is independent of language and industry, and sentiment is expressed through the opinions of the majority, not by polarity (+, -, neutral).
Building a semantic model
First we need to set up the semantic model for you. If we already have an engine built for your language or type of industry, onboarding is extremely easy. For specific business verticals, and more precise data interpretation, we are able to create new models - we can use any pre-generated feedback data you hold. The optimum amount of data is 20.000 feedbacks.
Reading topics on crunched data
The minimum feedback you need to collect to accurately utilize the semantic engine is around 1000 feedback for every batch. You can let the AI define feedback into topics by its own. You can also type specific drivers you are interested to track and see how they are reflected in the feedback you collected.