Semantic table structure identification in spreadsheets

Abstract

Spreadsheets are widely used in various business tasks, and contain amounts of valuable data. However, spreadsheet tables are usually organized in a semi-structured way, and contain complicated semantic structures, e.g., header types and relations among headers. Lack of documented semantic table structures, existing data analysis and error detection tools can hardly understand spreadsheet tables. Therefore, identifying semantic table structures in spreadsheet tables is of great importance, and can greatly promote various analysis tasks on spreadsheets. In this paper, we propose Tasi (Table structure identification) to automatically identify semantic table structures in spreadsheets. Based on the contents, styles, and spatial locations in table headers, Tasi adopts a multi-classifier to predict potential header types and relations, and then integrates all header types and relations into consistent semantic table structures. We further propose TasiError, to detect spreadsheet errors based on the identified semantic table structures by Tasi. Our experiments on real-world spreadsheets show that, Tasi can precisely identify semantic table structures in spreadsheets, and TasiError can detect real-world spreadsheet errors with higher precision (75.2%) and recall (82.9%) than existing approaches.