US Patent 11281567 Selecting an optimal subset of software tests for execution

For each software test of a plurality of available software tests, one or more code portions corresponding weight factor(s) of a first version of target codebase covered by the corresponding software test are identified. Information retrieval matching is utilized to calculate a corresponding match score for each software test of the plurality of available software tests based on comparisons of the identified one or more identifiers of one or more changed code portions of a second version of the target codebase with the one or more code portion identifiers and the one or more corresponding weight factors associated in the data structure with the corresponding software test. Based on the calculated match scores, one or more software tests of the plurality of available software tests are selected for execution to test the second version of the target codebase.