Our Engineering Department has something nifty brewing that should be ready next year…multilingual translation. It may not sound that enticing, but the concept of searching databases for new discoveries in different languages is a researcher’s dream! Using the Google Translation appliance, our multilingual translation feature will allow you to search from your native language, sending out your query to different publishers in different countries with different languages. Once results are retrieved, our Multilingual Translation feature will translate them back into your chosen language, aggregate the results, de-duplicate, then rank them, using our next-generation federated search software.
The opportunity for global information discovery by integrating this service in to our Explorit Research Accelerator federated search is monumental. Stay tuned for new developments with this feature.
Here’s our Press Release –
Deep Web Technologies Developing Multilingual Translator for Federated Search
SANTA FE, N.M., Sept. 3, 2009 — Deep Web Technologies is proud to announce development of a prototype of a multilingual translation capability for clients using its federated search applications. An early prototype of multilingual searching was demonstrated to the members of the WorldWideScience Alliance in June of 2009. This new feature, when fully developed and implemented, will translate a user’s search query into the native language of the collections being searched, will translate result titles and snippets back to the user’s original language and aggregate and rank these results according to relevance. The translation process will be seamless, making it simple to search collections in multiple languages from a single search box in the native language of the user.
“I’m very excited about this new multilingual translation capability. It will break down the language barrier for researchers in many countries, promoting global information discovery,” said Abe Lederman, President and CTO of Deep Web Technologies. “This will be particularly helpful in the areas of science, technology, government, and intelligence, where information may lie in hard-to-reach, non-English deep web collections.”
While many countries have a wealth of research content in local collections, much of this information has not been translated to other languages. Deep Web Technologies’ multilingual translation capability will augment the Explorit Research Accelerator(TM) federated search engine, allowing users to search multiple collections from different countries and languages simultaneously. The current prototype provides for searches of collections in English, Spanish, German, French, Portuguese, Russian and Chinese where the user selects the language desired for the results, which are aggregated, deduped, ranked and clustered in the selected language.
Deep Web Technologies’ Explorit Research Accelerator(TM) next generation federated search engine is focused on making the deep web accessible to all. It allows for comprehensive searching of many small, often unknown scientific, technical, government and subscription-based collections, as well as popular collections, all at the same time. Explorit displays results in an easy-to-follow, aggregated list, all from a single search request, providing researchers fast, efficient and comprehensive information discovery. Mainstream search engines, such as Google, Yahoo! and bing, which rank results on popularity instead of relevance, aren’t as strong for information discovery purposes.
“I foresee this multilingual translation capability as a major tool to speed the discovery process and disseminate critical scientific and technical discoveries, as well as time-sensitive government data,” mentioned Lederman. “Imagine finding relevant thesis and dissertations, journal articles and research reports from Chinese or Russian collections with a single, English-based search query.”
Deep Web Technologies is further improving its multilingual search capability and exploring with a number of our partners implementation of this capability on federated search-based portals such as ScienceResearch.com, WorldWideScience.org and the Defense Technical Information Center’s (DTIC) MultiSearch Portal.