Generally, computer training systems can only ask 'closed' questions: multiple choice, in effect. Although these can be 'dressed-up' to some extent, the 'pupil' is still being asked to choose from a limited number of pre-defined options, so the training effect is less than it might be.
Using our natural language understanding technology, we are adding 'open' questions to a well-known training system. This means that the course designer can include questions such as "What did the clerk do wrong in the last clip?", specify the correct answer or answers in natural language (eg "He hit the tax inspector over the head with a chair."), and the system will evaluate the student's free-text answer.