Communications of the ACM
Bible Helps AI Gain in Translation
Dartmouth College researchers trained an algorithm on different versions of the Bible; it now can convert written works into different styles without losing the context of the text.
Credit: Livemint
Dartmouth College researchers say the Bible can be used to improve computer-based text translation.
The team used the Bible's approximately 31,000 verses to produce 1.5 million unique pairings of source and target verses for machine learning training sets.
They developed an algorithm and trained it on different versions of the Bible, resulting in a system that converts written works into different styles for different audiences.
The team used 34 stylistically distinct versions varying in linguistic complexity, with texts fed into a statistical machine translation algorithm called "Moses," and a commonly used neural network framework called "Seq2Seq."
Said Dartmouth's Keith Carlson, "The English-language Bible comes in many different written styles, making it the perfect source text to work with for style translation."
From LiveMint
View Full Article
Abstracts Copyright © 2018 Information Inc., Bethesda, Maryland, USA
No entries found