Publications

Interactive Repair of Tables Extracted from PDF Documents on Mobile Devices

ACM Human Factors in Computing Systems (CHI)

Publication date: May 2, 2019

Jane Hoffswell, Leo Zhicheng Liu

Adobe Research thumbnail image

PDF documents often contain rich data tables that offer opportunities for dynamic re-appropriation and visualization. We describe a pipeline for extracting, analyzing, and visualizing PDF tables based on existing machine learning and rule-based techniques. Implementing and deploying this pipeline on a corpus of 447 documents with 1,171 tables results in only 11 tables that are correctly extracted and parsed for dynamic visualization. To improve the results of the automatic table analysis, we first present a taxonomy of errors that arise in the analysis pipeline and discuss the implications of cascading errors on the user experience. We then contribute a system with two lightweight interaction techniques (gesture and toolbar), for viewing, correcting, and visualizing PDF tables on mobile devices. In an evaluation with 17 users involving both a phone and a tablet, participants effectively repaired common errors in 10 tables, with an average time of about 2 minutes per table.

Learn More