Test Set Insights
In this section, you'll discover the importance of result analysis in testing processes and learn how to effectively interpret and utilize test outcomes. Each article delves into different aspects of result analysis, providing valuable insights, and identifying areas for improvement. Whether you're new to testing or looking to refine your approach, these resources will equip you with the tools and knowledge needed to get the most out of your result analysis.
-
What are Test Set Insights?
The Test Set Insights in Botium provide a rough estimation on the quality of your training data. A Static Analysis is done, which means that the data is not actually evaluated with an external chatbot engine like IBM Watson or Dialog...
-
LOOCV Insights
Botium simulates many nearly complete virtual NLP models, trained with all but one user example, and validates if the NLP model can extrapolate to the one missing user example. The simulation output for the sampled user examples are...
-
Humanification Insights
Botium simulates typical human behavior and typing habits, including typographic errors, by altering the training data and sending it to a virtual NLP model. The modified user examples and the simulation output are shown here. Where ...
-
Term Importance Insights
This heat map shows what terms are important for which intents. The darker the color, the more important the term is for the intent. Where to find Insights? You can find your Insights for each test in the Test Sets section. Navigate Us...