User acceptance test for software development in the agricultural domain using natural language processing

Software test case design is one of the most challenging activities since many actors with different backgrounds must cover most of the user's needs and expectations. In the agricultural domain tasks can be done in very different ways since practices vary worldwide. Thus, in this context, it is very hard to design test cases to validate requested functionality that automatizes some farm tasks. This paper proposes an approach to make the testing step easier, designing User Acceptance Tests (UATs) from requirements captured through scenarios. The scenarios capture the knowledge of different stakeholders (farmers) and using natural language processing tools, the approach proposed to consolidate the set of scenarios in a consistent and coherent base of knowledge organised in a tree, from where the design of test cases is extracted using the Task/Method model, a tool from the Artificial Intelligence.