...
By now (we hope) you have a functioning conversational agent! Congrats! We are sure by this time you have tried a few things out but we would like you to do some in-depth testing. We will have you use the agent for its designed purpose. You will have to use it a lot of times and try and vary the conversations you have. You should also try and trigger each pattern, page, intent, and filter to check if it all works. Part of your end report will be about how you did this(i.e. do you know how to trigger each intent, page, etc.). You will not only test your own bot though but also you will test a bot from another group and another group will test yours! More information can be found on the following pages about how to test each bot:Evaluation is an essential step to evolve your cooking assistant from an agent that can conduct a conversation, to an agent that many users can have a conversation with. Robustness is the keyword here. By testing, you can find out the deficiencies of the agent that need to be addressed so that it will not break down easily. This is typically done in a cyclical manner, where rounds of tests and improvements follow one another. You will follow two stages in the current project: system testing and a user study, which we will explain in further detail in the following subsections. The two stages will be reported on by your group in the end report.