$Exploration of Adaptive Random Test Replication Technology for Numerical Programs$

XJTU 2024 CSUC Submission14 Authors

31 Mar 2024 (modified: 03 Apr 2024)XJTU 2024 CSUC SubmissionEveryoneRevisionsBibTeXCC BY 4.0
Keywords: adaptive random test, replication technology, numerical programs, frontend and backend development
TL;DR: "It replicates various classical adaptive random testing algorithms and compares their effectiveness "
Abstract: $As a highly effective method for generating test cases, adaptive random testing is widely utilized across various fields, including fuzzing and AI testing. Among the numerous functional testing approaches, random testing serves as the fundamental method. It involves the random selection of test cases from the input domain until a specific condition is met, such as identifying errors within a software system. However, due to its limited utilization of additional information, the effectiveness of random testing is constrained. Consequently, adaptive random testing has been proposed to ensure the randomness of test cases and their even distribution throughout the entire input domain. This project focuses on numerical programs with an aim to replicate existing classical adaptive random testing algorithms and compare their efficacy. After studying the source code of the framework and conducting a thorough review of relevant academic literature, our team incorporated their own insights into the process of reconstructing the work of predecessors. Subsequently, we independently developed a comprehensive framework that facilitated customized data transmission, test case generation and execution, as well as evaluation procedures. Additionally, we utilized echarts to generate visually intuitive charts on the front-end.$
Submission Number: 14