As a technology company, one of the serious work with Google in the localization of language makes it easier to use user invests different countries. In Indonesia, Google promised to continue to deliver better services in the recognition of the Indonesian language.
Included in the service of the new Google Voice Search official in Jakarta, Wednesday (03/16/2011) is introduced. This is a voice-based search services where the user is saying what is sought, and will appear in Internet search results. recognize this machine is not always guaranteed to be correct 100 percent of the Indonesian language to learn.
Henk says: “Of course, as the user says the search terms is the debate in accordance with clear and immediate understanding of the results, depending on.” Prihatna, Indonesia Country Consultant, Google in Jakarta, Wednesday (03/16/2011). If you do not find a suitable sentence or phrase, the machine is a list of sentences that are similar recommendation. Once elected, the engine is Google to automatically learn new phrases. That’s how Google learn Indonesian.
However, before Google can learn before they had called the service to life already learned to recognize Indonesian words, sentence structure and grammar. In fact, exceptions are usually found in Bahasa Indonesia.
“Many challenges to locate the services in Indonesian,” he said. First, Indonesian Homograft writing the same word but different pronunciation and meaning.
Secondly, Indonesian and even after there EJT (Ajahn Enhanced) Dutch nor in many words. For example, the name of the Sukarno and Soekarno the same pronunciation but different fonts. should seek voice in a position to overcome them.
Third, some people use the phrase mixture Indonesia Indonesian and English. Services must be able to identify words in different languages in one sentence.
Fourth, there is currently no standardization of the pronunciation of words in Bahasa Indonesian. EYD to write new, but hangs around the debate of access, so that pronunciation varies on ethnic origin or region of origin of the speakers.
“To recruit Google since last year and to provide time for the engineer, the collected data,” said Heng. He said, is a way to involve as many as 500 student volunteers. They were asked to pronounce the word list so that the database collected by the hundreds of thousands of phrases. This database will then be in a server cloud that will translate the text to search stored.
According to the story Heng directly involved in the project, he also sent to a Google research center in New York to improve the model voice search in Indonesian language. Machine try to do a search with voice and over again until the level of truth as possible. And over time, machine learning new sets of many users of the service.