一句话：提出基于特征聚类与簇细化的电子表格缺陷检测工具 SGuard，在精度与召回间取得平衡并超越现有技术。
电子表格被广泛使用，但容易包含各类缺陷。本文提出缺陷检测工具 SGuard，通过学习电子表格特征，将具有相似计算语义的单元格聚类，再对聚类进行细化，从而把异常单元格识别为缺陷。该方法在缺陷检测中达到87.8%的精度和71.9%的召回率，F-measure为0.79，优于现有电子表格缺陷检测技术。作者还介绍了 SGuard 的实现与使用方式，并提供视频演示和公开下载仓库。
Spreadsheets are widely used but subject to various defects. In this paper, we present SGuard to effectively detect spreadsheet defects. SGuard learns spreadsheet features to cluster cells with similar computational semantics, and then refines these clusters to recognize anomalous cells as defects. SGuard well balances the trade-off between the precision (87.8%) and recall rate (71.9%) in the defect detection, and achieves an F-measure of 0.79, exceeding existing spreadsheet defect detection techniques. We introduce the SGuard implementation and its usage by a video presentation (https://youtu.be/gNPmMvQVf5Q), and provide its public download repository (https://github.com/sheetguard/sguard).
DOI 原文 ·
@inproceedings{paperbot5339,
title = {SGUARD: A Feature-Based Clustering Tool for Effective Spreadsheet Defect Detection},
author = {Da Li and Huiyan Wang and Chang Xu and Ruiqing Zhang and Shing-Chi Cheung and Xiaoxing Ma},
booktitle = {ASE},
year = {2019},
doi = {10.1109/ase.2019.00122}
}