We are now on the 2nd and final day of the UKSTAR talks. Thankfully, I’d been able to get a decent nights sleep despite attending networking evening that had been arranged at a local pub. I arrived nice and early, in time for breakfast and the Lean Coffee event at the huddle area. This was followed by another day of brilliant talks.
Keynote 4* – Why do we automate?
This is a brilliant question that many neglect to ask. In this talk, Bas discusses the common mistakes that occur in test automation such as attempting to automate everything. Instead, we should be asking questions about why we want to automate so we are sure that we are doing the right thing. It is not a case of one size fits all, the approach to test automation will be different for each project. Automation is wonderful but only if done right.
It is important to, every now and then, step back and ask “why?”
- Why do we need automation?
- Why this tool?
- Why these tests?
- Why this percentage of tests?
Only with a satisfactory answer should we proceed.
Bas used the following quote from the movie Jurassic Park to summarise his point:
“Your scientists were so preoccupied with whether or not they could, they didn’t stop to think if they should.”Ian Malcolm, Jurassic Park
Bas also used another brilliant analogy: Test automation is like bubble wrap. Its fun to play with but it has its limits and can give a false sense of security. He discusses this analogy more here.
*This was originally scheduled to take place at the end of the day not the beginning. I’ve labelled this as keynote 4 to match the program even though it took place before keynote 3.
Cognitive Testing Talks
Cognitive Testing – Insight Driven Testing Powered By ML by Poornima Shamareddy
Cognitive QA by Wayne Rutter
The next 2 talks were on very similar topics from the modern tester track.
The first talk, Cognitive Testing – insight driven testing powered by ML by Poornima Shamareddy, talked about the development of a self learning system that used data mining and pattern recognition to rank and prioritize test cases. During the talk, Poornima walked us through the process of developing the application including the benefits achieved.
The second talk, Cognitive QA by Wayne Rutter, discussed investigation into ways to identify the amount of test resource required for each area of an application. Wayne went into great detail about some of the different machine learning methods including supervised learning, such as using classification methods, and unsupervised learning, such as clustering. This talk was especially impressive as Wayne was part of the speakeasy program having never given the talk before.
Artificial Intelligence and machine learning concepts are something I’ve not covered since university. My memory of the subjects is a little vague but both talks provided enough basic information about machine learning techniques to make the talk informative to anyone regardless of prior knowledge. It was interesting to see examples of practical applications of artificial intelligence, natural language process and machine learning from a test perspective.