Web Exclusive

April 23, 2002

Listen Up

Is "audio mining" the next frontier for business intelligence?

By Jeanette Perez

Valuable information about your customers and product can be found in the spoken word. However, the dilemma becomes how to search and extract information from speech. Audio mining technology may be the answer. Just as data mining software searches and organizes written data, audio mining can index, search, and extract data from phone calls, speeches, media broadcasts and so on, and many companies are using it as the next strategic business application.

"In effect, audio mining engines create a high-level structured summarization of the spoken language that can be searched and browsed using different criteria Ð such as the names of people, places, organizations, and topics mentioned in the transcript as well as the identities and locations of the speakers in the recording," said Curt Hall, senior consultant at Cutter Consortium.

The ability to search recorded voices can prove very useful in the fight for national security. As Homeland Defense Secretary Tom Ridge looks to technology as a means of keeping the country safe, he may want to check into audio mining. ScanSoft Inc., a supplier of imaging, speech, and language solutions, has an audio mining product already being used for security in a prison.

"[The prison] is using [ScanSoft's audio mining technology] to manage the recordings of prisoners phone calls. The product is able to sort out and prioritize the conversation and then focus attention on the conversations that have flagged words," said Robert Weideman, vice president of worldwide marketing at ScanSoft. Weideman believes that audio mining will definitely eventually be used for other types of security as well, but cautioned that privacy may be an issue. "Of course there is always the issue of privacy. But where recording information is allowed, as with the inmates recorded conversations, it will definitely be used."

According to a Cutter Consortium article, audio mining can also be used for strategic advantage in call centers and ScanSoft says that broadcast media companies, such as Radio Free Asia are already using Dragon MediaIndexer, their speech indexing technology. Call center operators can directly refer to recorded conversations, giving better customer service. Broadcast media companies are able to quickly refer back to earlier broadcasts when new stories are breaking and find audiovisual information.

"There is a wealth of valuable content that is trapped in the world's audio and visual files," said Jackie Fenn, vice president and research fellow at Gartner Group. "Speech indexing and audio mining is an emerging application of speech recognition that will open up this content to corporate, academic, and government customers."

• Privacy
• Your California Privacy Rights
• Copyright © 2008 United Business Media LLC