nlrpBENCH is a set of tasks, grouped into benchmarks. A task is a NL requirements document and possible solutions. A task is associated with a task type. At the moment there are two task types: model extraction and text correction.
Every task has an expected result and for every task type there are metrics which determine the quality of a solution (recall, precision, and F-measure).
Additional task types will be added (e.g. test code generation).
Pages in category "Benchmarks"
The following 2 pages are in this category, out of 2 total.