Determining the quality of my test cases

[W Umeed]

Hello

I am using Jira XRAY Atlassian product as my test management suit. We have a number of automated and manual test cases and I am working on an AI model to evaluate the quality of my test cases. 

My problem is the evaluation of the readability and completeness of the test steps in a test case. Ideally, I am looking for a dataset of POS instructional tagging, based on which I can train my model and then using nltk I can identify sentences starting with an actionable verb, and eliminate any probablistic terms like mostly, probably etc from my test steps. 

Is there anything similar already implemented? OR do we have a dataset based on which we can benchmark the quality of our test steps in a test?

[Manuel]

Hi @W Umeed ,

I'm not aware of any tool for that specific purpose but perhaps there is something.

However, one thing is "quality of the writing".. another thing is "quality of the actual tests" considering what they target and if that is relevant or not. That is very hard to achieve as you need context about what's relevant on a given moment, in the context of your project.

I would move away from investing on quality of the writing and consider more improving the skills and understanding of testing and what really matters.. and that requires collaboration, domain understanding, and putting your "hands dirty" :)