Analysis of the effectiveness of software testing technologies for information systems

Abstract

The purpose of this work is to analyze the effectiveness of software testing. Modern approaches to software testing are analyzed, and their advantages, limitations, and areas of practical application are identified. To conduct a computational experiment, an experimental dataset in CSV format was created, containing approximately 15,000 records that characterize the testing parameters, specifically costs, execution time, and the number of defects found. Models were constructed to evaluate the economic efficiency of various testing types, enabling the determination of time limits for return on investment. Based on the results of the experiment, practical recommendations were formulated for choosing the optimal type of testing: manual testing is appropriate for small projects due to low costs, automated testing is appropriate for medium-sized projects with repetitive scenarios, and intelligent testing is suitable for large and dynamic systems where adaptability and high accuracy in detecting defects are critical.