BIROn - Birkbeck Institutional Research Online

    Enhancing test reuse with GUI events deduplication and adaptive semantic matching

    Liu, S. and Zhou, Y. and Ji, L. and Han, Tingting and Chen, Taolue (2024) Enhancing test reuse with GUI events deduplication and adaptive semantic matching. Science of Computer Programming 232 (103052), ISSN 0167-6423.

    [img] Text
    1-s2.0-S016764232300134X-main.pdf - Published Version of Record
    Restricted to Repository staff only

    Download (3MB)

    Abstract

    Developers typically employ Graphical User Interface (GUI) testing to ensure the expected behavior of applications, but they face the challenge of designing appropriate test cases with functional features. Recently, researchers have proposed several test reuse methods based on semantic matching to alleviate the burden. However, the limited text semantic information and semantic matching rules between events severely limit the existing test reuse methods. In this paper, we propose TREADROID (Test Reuse EnhAncer for anDROID applications), a framework that combines GUI events deduplication with the adaptive semantic matching strategy to enhance the usability of the reused tests. Considering the connection between widget attribute texts, we categorize attributes and measure widget similarity based on the same corresponding attributes as well as across attributes in the same group. In addition, we propose a deduplication strategy for GUI events to reduce the redundancy caused by reusing a test with unique functionality. To further bridge the semantic gap, we design a two-stage adaptive matching strategy to search for the target test with functionality closer to that of the source test. Experimental evaluation against the baseline methods on 25 applications demonstrates that: (i) the adaptive semantic matching strategy overall improves the performance of widget mapping; (ii) GUI events deduplication dramatically increases the precision of events on average, even reaching 100% for multiple tests; (iii) TREADROID can significantly reduce the manual effort of creating tests for similar applications.

    Metadata

    Item Type: Article
    School: Birkbeck Faculties and Schools > Faculty of Science > School of Computing and Mathematical Sciences
    Depositing User: Tingting Han
    Date Deposited: 18 Mar 2024 15:40
    Last Modified: 18 Mar 2024 19:24
    URI: https://eprints.bbk.ac.uk/id/eprint/53236

    Statistics

    Activity Overview
    6 month trend
    2Downloads
    6 month trend
    62Hits

    Additional statistics are available via IRStats2.

    Archive Staff Only (login required)

    Edit/View Item Edit/View Item