Test Intelligence: Find more errors in less time based on your own data
Many teams need to test more and more functionality in less and less time. Historically grown test suites often reach their limits because they test too much and too little at the same time. Too many, since they contain tests that generate costs but offer little added value compared to similar tests. Not enough, since important functionality remains untested.
In the lecture we present analysis techniques to show these problems in your own system. It covers: version control system history analysis, test gap analysis, Pareto test analysis, and test impact analysis. For each analysis, we show the underlying concepts from research and our experience in practical use.
Assisted Requirements Engineering - what will remain in the hands of the future Requirements Engineer?
Requirements engineering today is widely considered one of the most difficult and risky activities in software and systems engineering. With a daily evolving field of natural language processing (NLP), the question is: Which tasks will - independent from any technological and methodological advancements - stay in the hands of the requirements engineer and which tasks will be automated? This talk will take a biased look into the crystal ball. Based on analogies from programming and autonomous driving, and based on an analysis of the abilities of NLP and text analytics, I present a biased and speculative vision of the life of a future requirements engineer.
Am I beautiful? Body Mass Index for requirements
Useful and less useful key figures from the precise measurement of more than 130 requirement specifications
For years we have been measuring requirement specifications from a wide variety of industries, projects and contexts in the course of our consulting.
In this session we would like to present a small selection of the results. The lecture will be a colorful smorgasbord of interesting, useful, worth knowing and surprising numbers, data and facts from the world of requirements. You'll hear what metrics can be used to instantly tell if the spec is well-structured, how long the average spec is, what the most popular terms are, how long sentences should be and how long they actually are on average, and how much technical debt is in the average specification.