Kanjingo – a mobile app for post-editing

Kanjingo – A Mobile App for Post-Editing Sharon O’Brien CNGL/SALIS Dublin City University Ireland sharon.obrien@dcu.ie Joss Moorkens CNGL/SALIS Dublin City University Ireland joss.moorkens@dcu.ie Joris Vreeke CNGL Dublin City University Ireland joris.vreeke@dcu.ie Abstract This paper describes the Kanjingo post- editing application for smartphones. The application was developed using an agile methodology at the Centre for Global Intelligent Content (CNGL) at DCU and a first stage of user testing was conducted using content from Translators Without Borders.1 Initial feedback on this app was quite positive. Users identified some particular challenges, e.g. input and sensitivity limitations, insufficient Help, lack of automatic punctuation and capitalization. Development and further testing are ongoing and may include interactive MT, speech as input and focus on Asian languages as target languages in the future. 1 Introduction Kanjingo is a mobile app for translating a source text and post-editing machine translated target text on a mobile interface. It was developed in the CNGL (Centre for Global Intelligent Content) at Dublin City University.2 This paper describes the first round of user testing where the objective was to obtain feedback and improve the application. 2 User Testing 2.1 Motivation The objective of the first stage of user testing focused on Kanjingo‟s suitability for post-editing © 2014 The authors. This article is licensed under a Creative Commons 3.0 licence, no derivative works, attribution, CC-BY-ND. 1 We are very grateful to Translators Without Borders for their collaboration in this project. 2 This research is supported by the Science Foundation Ireland (Grant 12/CE/I2267) as part of CNGL (www.cngl.ie) at Dublin City University. machine translated output in a mobile scenario. The motivation for doing so is based on the increasing evidence that volunteers are willing to translate or post-edit for causes they wish to support (Munro, 2010; Petras, 2011) Our use case scenario for this first round of testing is volunteers for an organization such as “Translators Without Borders” (TWB). The volunteers wish to contribute to the translation effort of this organization, but possibly only have time to translate or post-edit a few segments of text per day on their way to and from work. Our assumption is that volunteers may not wish to sit at a desk to do this work and might like to post- edit a few segments of text while waiting at a bus stop, for example. The Kanjingo App is not intended to replace a desktop CAT environment. However, since MT suggestions sometimes need to be deleted outright due to poor quality and retranslated by a human, we decided to also test the App‟s potential to support the human translation task in addition to the post-editing task. Figure 1. The Kanjingo post-editing screen 137 2.2 UI Description When the App is first accessed, the user selects a language pair, e.g. English-French. Source segments are listed in the initial screen presented to users. The user selects a source segment at which point a machine translated segment is presented on the screen in a vertical tiled format (see Figure 1). If at …

Content truncated to 500 words