Statistical Glossing - Language Independent Analysis in Bible Translation

摘要

Bible translation sets a number of particular challenges for machine translation and analysis. The nature of the work is such that translators are often working with local vernacular languages for which there are little in the way of lexica and other linguistic databases. Commercial text processing and translation systems are rarely able to contribute. Translation Consultants (TCs) charged with advising translation teams often have few computer based aids to assist them in their work. This paper describes the development of the Statistical Glossing Tool (SGT) which ships with the Paratext translation editor. SGT is a language independent method for the analysis of bible translations. It provides an objective assessment which TCs can use to help them identify key issues in a new translation where further work may be needed. SGT requires no information about target languages other than the text itself.