Exact and Near-miss Clone Detection in Spreadsheets

Spreadsheets are used extensively in business, in many domains. The applicability of software engineering methods to spreadsheets has been a topic of research for several years, but the main focus has been on analyzing the formulas, and not on analyzing the data in the spreadsheets. One of the factors that plays a role in spreadsheet data quality is the occurrence of clones in the spreadsheet data. Clones in data are caused by copy-pasting. This is a very common practice in spreadsheet use, however, it can have a negative impact on the spreadsheet's quality, since 1) editing the copied data needs to be done in multiple places increasing maintenance effort and 2) when editing, some copies might be forgotten, leading to errors. Clone detection has been proven useful in the realm of source code analysis, in two different forms: exact clones, and clones that differ slightly, called near-miss clones. Because of the success of clone detection and removal in source code, it seems feasible to research the applicability of both techniques on clones in spreadsheet data. Our work shows that this is a promising avenue.}电子表格（Spreadsheet）在商业及众多领域中得到了广泛应用。多年来，将软件工程方法应用于电子表格的研究一直是学界关注的课题，但既往研究的核心关注点多集中于电子表格的公式分析，而非其内部的数据内容。影响电子表格数据质量的关键因素之一，是电子表格数据中存在的数据克隆（Clone）现象。数据克隆通常由复制粘贴操作所引发，尽管该操作在电子表格的使用场景中极为普遍，但会对电子表格的数据质量产生负面影响：其一，需在多处对复制得到的数据进行同步编辑，大幅增加了维护成本；其二，编辑过程中可能遗漏部分复制内容，进而引发各类错误。代码克隆检测已被证实可有效应用于源代码分析领域，其检测类型主要分为两类：精确克隆（Exact Clone）与近似克隆（Near-miss Clone，即存在细微差异的克隆）。鉴于代码克隆检测与移除技术在源代码领域已取得成熟应用，将这两类技术拓展至电子表格数据克隆的研究具备可行性，本研究证实该方向具有可观的发展前景。