KR20010064061A - Search Engine with Voice Recognition - Google Patents

Abstract

PURPOSE: A search engine having a voice recognizing function is provided for a user to search data without using a keyboard or a mouse operation by converting a voice signal into characters and searching video/radio data in a database based on the contents of the characters and outputting video/radio in accordance with a search result when the user inputs a voice with respect to search contents to a client computer. CONSTITUTION: An AGC circuit(22) converts a voice signal inputted by a mike(21) from the exterior into an electric signal and adjusts a size of the converted electric signal. A voice input unit(20) converts the analogue electric signal of the AGC circuit(22) into a digital signal and inputs the signal to a client computer(10). The client computer(10) transmits the voice signal to a server computer(80) and receives search result data transmitted from the server and outputs video/radio data. A video output unit(40) and a radio output unit(30) process the output video signal and the output radio signal of the client computer(10) and output the signals to a monitor screen(41) and a speaker(31). A data storing unit(70) stores background video data and sound data with respect to various search data. The server computer(80) includes a voice recognition unit storing the various search data in the data storing unit(70), analyzing the audio signal from the client computer(10), recognizing the contents corresponded to the audio signal and converting the contents into character data. In addition, the server computer(80) searches data corresponded to the requested contents and outputs the data to the client computer(10).

Description

Search engine with speech recognition {Search Engine with Voice Recognition}

According to the present invention, a voice is inputted to a client (or user) computer to search data by voice, a voice signal is recognized by a server to analyze a user's requested search, and a search result is output to the client. A search engine having a function.

The conventional search engine selects a search portion using a keyboard or a text input method using a keyboard, and thus performs a search by receiving selected data. Therefore, since the user requires to manually input a sentence or click with a mouse, it is inconvenient for a user with a physical disability because the user must always use the keyboard.

Accordingly, the present invention has been made to solve the above problems, the present invention, when the user voice input the search content to the client computer, the server converts the voice signal to the text using the speech recognition means and the text content By analyzing and searching video and audio data corresponding to the search request from the database and sending it to the client computer, the client computer can output video and audio according to the search result, so that the user does not use the keyboard or mouse. The goal is to enable data retrieval without the need.

In order to achieve the above object, in one embodiment of the present invention, voice input means for inputting a user's voice into a client (user side) computer; Voice recognition means for analyzing a voice signal received through the voice input means and converting the voice signal into a word corresponding to the voice signal; Content analysis means for analyzing the input text content recognized by the speech recognition means; Retrieving means for performing data retrieval according to the content analyzed by the content analyzing means; The present invention provides a search engine having a voice recognition search function, comprising video and audio output means for signal-processing information on a searched data, a video signal, and an audio signal, respectively, and outputting them to a monitor screen and a speaker of a client.

1 is a block diagram of a search engine having a speech recognition function according to an embodiment of the present invention;

2 is an operation flowchart of a search engine having a voice recognition function according to the present invention;

3 is a block diagram illustrating an exemplary embodiment of a client and a server in a network configuration according to the present invention.

<Explanation of symbols for the main parts of the drawings>

10: client (user side) computer 20: voice input unit

21: microphone 22: AGC circuit

30: audio output unit 31: speaker

40: video output unit 41: monitor

70: data storage 80: server (host) computer

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1 is a block diagram schematically showing the configuration of a search engine having a voice recognition function according to the present invention. The voice signal input using the microphone 21 from the outside is converted into an electric signal, and the converted electric signal. AGC circuit 22 for adjusting the size of the audio signal, and a voice input unit 20 for converting and inputting an analog electric signal output from the AGC circuit 22 into a digital signal to a client (user side) computer, and through the voice input unit. A client computer 10 for transmitting an audio signal from the outside to the server (or host) computer 80, receiving the search result data sent from the server, and outputting video and audio data; and outputting the client computer 10 The video output unit 40 and the audio output unit 30 to process the video signal and the output audio signal, respectively, and output them to the monitor 41 screen and the speaker 31, and various tests. Data storage unit 70 for storing the background image data or sound data according to the color data and the content of the search data, and stores various search data in the data storage unit 70 and analyzes the audio signal sent from the client 10 computer And voice recognition means for recognizing a content corresponding to the voice signal and converting the text data into text data, and searching for data corresponding to the requested content to read a desired type of content from the stored data. Server (or host) computer 80 to output.

The client computer 10 receives a signal from the voice input unit 20 that receives a human voice and converts it into an electrical signal, and converts the signal into a matching word based on the pronunciation and accent of the user. It may also be configured to include a voice recognition device 20 for outputting a text to the).

The voice input unit 20 may include an AGC circuit 22 for adjusting the output signal gain of the microphone 21 for receiving a human voice and converting it into an electrical signal, and the signal output from the microphone 21. It can also be configured as a speech recognition device 20 that receives the input to convert the word into a matching word based on the pronunciation and accent of the user to output to the client computer.

The server computer 80 converts the voice signals sent from the plurality of client computers 10 into text using speech recognition means, and processes the data corresponding to the text contents to search for the corresponding client computers. You can also send.

The data storage unit 70 stores search data and video and audio data about the data.

2 is an operation flowchart of a search engine having a speech recognition function according to the present invention, and shows a signal processing flow executed by the server computer 80.

As shown in the figure, when the client computer 10 connects to the server computer 80, the server computer 80 waits for input of a voice signal from the client computer 10. When the voice signal is input, the voice recognition means detects a character corresponding to the voice pattern and performs character conversion.

In the drawing, the client computer 10 constructs and outputs an initial screen for searching at the same time as the client computer 10 is connected to the server computer 80, and waits for a voice search request through the initial screen ( S411); A second step (S412) of converting a voice signal sent from the client computer 10 into a text by executing a voice recognition means subroutine for voice recognition in the server computer 80 when a search request is made by a user; ; A third step (S413) of executing content analysis means to analyze input text content recognized by the speech recognition means; A fourth step (S414) of executing a search means to perform a data search according to the content analyzed by the content analysis means; The fifth step (S415) may be performed by executing data output means to process the information on the searched data, the video signal and the audio signal, respectively, and output the video and audio output to the monitor screen and the speaker of the client.

In the first step, for example, a search type that can be classified by "news", "Software" or "chatting", and a step of outputting a guide screen for the user to select according to each search type; The method may further include waiting for a user to input a voice through the guide screen.

The second step may further include outputting a guide screen requesting re-speech input when a word that cannot be recognized is input by executing voice recognition means for speech recognition.

In addition, the third step further includes the step of executing the content analysis means to investigate the type and grade of the search, and outputs a warning message for notifying that the search is not a correct search request if the search is not a permitted level (S416). can do.

In the fourth step, if it is determined that the content corresponding to the content of the data requested by the user is not found in the database from the result of performing the content analysis means, an operation of outputting a warning message for notifying that the information is not a correct search request is performed (S416). It may further include. At this time, the warning message or re-entry request information may be output as a text, or voice or text may be used at the same time.

3 is a block diagram illustrating an exemplary embodiment of a client and a server in a network configuration according to the present invention. The server computer 600 and the plurality of client computers having the components shown in FIG. And 610-630. The plurality of client computers 610 to 630 are connected in parallel to the server computer 600 through a communication line to exchange data and to perform a one-to-many search.

The above processes are described as specific examples.

First, the process of the user searching the Korean Patent Office data using a computer proceeds as follows. First, a user speaks into the microphone 21

"Korea Patent Office"

The client computer 10 transmits the voice signal to the server computer 80 via a communication line. The server computer 80 converts the input voice signal into a character, retrieves the data corresponding to the recognized character "Korea Patent Office" in the stored data storage unit 70, and reads the voice data and image data related to the Korea Patent Office homepage. It transmits to the computer 10 using a communication line. The client computer 10 outputs the received video and audio data to the monitor 41 and the speaker 31 of the user through the video output unit and the audio output unit.

In particular, the present invention enables data retrieval by voice command at any position where voice transmission is possible by using voice recognition means that receives voice and converts the voice into words. Accordingly, the present invention has the advantage that even a user with a physical disability can search for a material using voice, and also easily surf the Internet or home shopping.

Claims (10)

Voice input means for inputting a user's voice to the client (user side) computer; Voice recognition means for analyzing the input voice signal and converting the contents of the voice signal into text; Content analysis means for analyzing the input text content recognized by the speech recognition means; Retrieving means for performing data retrieval according to the content analyzed by the content analyzing means; And a video and audio output means for signal-processing the information on the searched data and the video and audio signals, respectively, and outputting them to the monitor screen and the speaker of the client. The method of claim 1, wherein the output voice signal of the voice input means of the client (user side) computer is transmitted to the server, The server includes a voice recognition means for analyzing the voice signal transmitted from the server, and converting the content meaning of the voice signal into a text, and content analysis means for analyzing the input text content recognized by the voice recognition means. Search engine with voice recognition search. The client (user side) computer according to claim 1, wherein the client (user side) computer includes voice recognition means for analyzing the output voice signal of the voice input means of the client (user side) computer and converting the content of the voice signal to text. And a content analysis means for transmitting the output text of the means to the server and analyzing the content of the text received at the server. 2. The method according to claim 1, wherein the output voice signal of the voice input means of the client (user side) computer is analyzed, and the contents of the output character of the voice recognition means and the voice recognition means for converting the content of the voice signal means into text are analyzed. And a retrieval means for including the content analysis means in a client (user side) computer, transmitting the output contents of the content analysis means to the server, and retrieving the content received from the server. Search engine. The method of claim 1, further comprising: performing a rating analysis on the data information retrieved from the server, deleting search contents other than the rating set by the user, and processing the information, the video signal, and the audio signal included in the rating range, respectively. Search engine with a speech recognition search function further comprises a video and audio output means for outputting the monitor screen and the speaker. Voice input means for inputting a user's voice to the client (user side) computer; Voice recognition means for analyzing the input voice signal and converting the contents of the voice signal into text; Content analysis means for analyzing the input text content recognized by the speech recognition means; Web site search means for performing a web site data search according to the content analyzed by the content analysis means; Data output means for processing information on the retrieved web site data and outputting the information to a monitor screen and a speaker of the client; Web site retrieval means having a voice recognition retrieval function comprising video and audio output means for signal-processing the video signal and the audio signal of the retrieved web site data and outputting them to the monitor screen and the speaker of the client. 7. The method according to claim 6, wherein the output voice signal of the voice input means of the client (user side) computer is transmitted to the server, The server includes a voice recognition means for analyzing the voice signal transmitted to the server, and converting the content meaning of the voice signal into a text, and content analysis means for analyzing the input text content recognized by the voice recognition means. Web site search engine with voice recognition search. The client (user side) computer according to claim 6, wherein the client (user side) computer includes voice recognition means for analyzing the output voice signal of the voice input means of the client (user side) computer and converting the content of the voice signal to text. And a content analysis means for transmitting the output text of the means to the server and analyzing the content of the text received by the server in the server. 7. The method according to claim 6, wherein the output voice signal of the voice input means of the client (user side) computer is analyzed, and the contents of the output character of the voice recognition means and the voice recognition means for converting the content of the voice signal means into text are analyzed. And a web site searching means for including the content analyzing means in a client (user side) computer, sending the output contents of the content analyzing means to the server, and performing a web site search of the content received in the server. Website search engine with speech recognition search. The client of claim 6, further comprising: performing a rating analysis on the website information retrieved from the server, deleting search contents other than the rating set by the user, and processing the information, the video signal, and the audio signal included in the rating range, respectively. Web site search engine with voice recognition Internet search function, characterized in that it further comprises a video and audio output means for output to the monitor screen and the speaker.