UMD Team Awarded $14.4M to Build System That Can Translate Any Language
Universal translators that make any language understandable are common in science fiction, but now Maryland researchers are involved in a project to make the concept a reality.
The four-year initiative, funded by a $14.4 million grant from the Intelligence Advanced Research Projects Activity, aims to produce a language processing system that allows a user to type in a query in English and have information returned in English—even if the content is available only in a lesser-known languages or dialects.
The U.S. intelligence community-funded research has clear national security implications, but the UMD team’s leader says it could enhance communications for people around the globe.
“Today’s internet brings us closer together than ever before, but the diversity and richness of human language remains a challenge,” says Douglas Oard, a professor in the College of Information Studies.
The interdisciplinary research includes experts in natural language processing, speech processing and information retrieval from Maryland, Columbia University, Yale University, the University of Cambridge and the University of Edinburgh.
Joining Oard at Maryland are Philip Resnik, a professor of linguistics; Marine Carpuat, assistant professor of computer science; and Hal Daumé, professor of computer science. All four have joint appointments in Institute for Advanced Computing Studies.
The system they’re building, called SCRIPTS—System for Cross Language Information Processing, Translation and Summarization—will take advantage of the latest computing advances, including machine-learning algorithms that can sift through large amounts of human language, looking for commonalities in syntax and semantics.
The goal is artificially intelligent systems able to teach themselves the wide world of languages, rather than language-by-language training of systems, Oard says.
“Computers can be trained to transform human language in many useful ways, but today that training process is still too expensive to affordably be applied to all the world’s languages, and too dependent on the artisanal skills of a small number of experts,” he says.
When completed, SCRIPTS will be able to transcribe speech from multiple sources such as videos, news broadcasts and some types of social media. It will also process text documents like newspapers, reports and social media posts.
The system will use multiple strategies, such as matching an English query against translated documents, then summarizing the result, as well as searching and summarizing directly in the foreign language, then translating the selected summaries into English.
One of SCRIPTS’ specialties will be languages such as Tagalog or Swahili, which although widely spoken, have small digital footprints. Known as “low-resource languages,” these generally can’t be translated by computer.
This is where new technology will come into play. Deep learning-based translation systems under development at Maryland will take limited amounts of information from the low-resource languages, churn it with other language-related data from better-resourced languages, and come up with powerful new tools that will allow for the manipulation and transformation of content in those languages.
“In order for us to be able to do this kind of work, we need the ability to build new computing infrastructures that weren’t the same ones people were using as recently as five years ago,” says Carpuat, an expert in multilingual text analysis.
Perhaps of greatest significance, the researchers say, is that SCRIPTS is designed to incorporate four key areas of language processing—speech recognition, machine translation, cross-language retrieval, and information summarization—into one, robust platform.
“(A)ll needed to be done within separate systems,” says Resnik, a computational linguist. “Now—with the use of deep learning neural networks—it allows us to combine functions and do a single ‘training’ of the system across multiple functions quickly and efficiently.”