Abstract
Digital humanities and translation scholars utilize off-the-shelf tools to align multiple related translations. These tools generally rely solely on domain expert knowledge and do not exploit the recent advancements in computational linguistics and text mining. This paper presents AlignVis, a visual tool that provides a semi-automatic alignment framework to align multiple translations. It presents the results of using text similarity measurements and enables the user to create, verify, and edit alignments using a novel visual interface. The design consists of three main components: the alignment editor canvas, the post-edit area, and the user options panel. AlignVis exploits both close and distant reading and is designed to help digital humanities and translation scholars enhance the process of text alignment for multiple translations. The design of AlignVis is driven by iterative discussions with the domain expert which resulted in five benefits: presenting an overview of the aligned translations, support for multiple alignments, enhancement and acceleration of the alignment process, alignment refinement, and testing different similarity measurements. We evaluate AlignVis with domain expert feedback and a comparison with a standard alignment tool and computational and visual alignment tools.