Voice search

Voice search is a speech recognition technology that allows you to translate a user's voice request into a text form, which is then transmitted to a standard database search system.

Applications most explicitly associated with the term “voice search” are based on the use of speech recognition systems and often speech synthesis for the automated return of search results. Voice search is implemented in the following areas:

• alphabetical directory, search for a company by name or category, search for a person in the list;
• search for information, such as news, finance, traffic jams, driving directions, weather or information on movie theaters (this often uses the management of multi-level voice menus);
• search on the Internet, similar to if the request was printed in the search line, but carried out by voice (sometimes leading to text output on a mobile phone);
• selecting options from a long list of mobile services, on a portable device or in the car system (songs or radio stations).

All these search forms are initialized by voice query, the result of which is either the choice of the desired action or the delivery of information in the form of synthesized speech. A distinctive feature of these core applications is quick access to information, minimizing its efforts to find it.

The reason for using voice is often that the device you are using is a mobile phone, where text input is inconvenient or dangerously distracting, and where it is difficult to view long lists due to the small screen.

Another category of applications that can be considered part of voice search is “voice dialing” - search for a contact in the directory. The service consists in the fact that after a voice call to the system and determination of the search name, it makes a call to the selected contact.

Sometimes, the term “sound search” means a search by audio (or audio track of a video fragment) to find speech fragments and convert them into searchable content. Applications include finding in a video or podcast a specific topic or company mention; radio companies can convert archived audio / video content into a more user-friendly form, and broadcasters can more easily find information. This technology is based on speech technologies that allow you to analyze the voice content of files.

When using these technologies in call centers, they are called “speech analytics”, which complements the recognized speech information in order to do some statistical and language analysis using ready-made textual and phonetic material. Such applications allow you to search for specific sound fragments by the necessary attributes (for example, determine dissatisfied reviews of called customers).

Converting voice files to text for further text search by them exists, for example, in such services that allow you to convert voice mail to text for easier searching and viewing, and also allow further forwarding of voice mail in the form of e-mail or sms. Services are also available that provide the ability to leave voice notes on the phone and convert them to text. One of the main goals of this service is to make voice content easily archived and searchable.