KR20190036463A - QUERY AND RESPONSE SYSTEM AND METHOD IN MPEG IoMT ENVIRONMENT - Google Patents

Abstract

The present invention relates to a system for query and response in a moving picture experts group (MPEG) internet of media things (IoMT) environment. The system comprises: an IoT terminal receiving and transmitting articulation information and receiving and providing query and response result information; and an articulation analyzing server performing articulation analysis on the articulation information provided from the IoT terminal according to MPEG IoMT data format and providing the query and response result information after using the analyzed articulation information to perform query and response with a query and response server.

Description

TECHNICAL FIELD The present invention relates to a QUERY AND RESPONSE SYSTEM AND METHOD IN MPEG IOMT ENVIRONMENT,

The present invention relates to an apparatus and method for accurately detecting a correct answer desired by a questioner about device operation, information transfer, and query response in response to various demands of a user for implementing a query response system and method in an MPEG IoMT environment will be.

The conventional question-and-answer technique has difficulty in solving the needs of various users because it depends on only the question sentence inputted by the questioner and finds the correct answer.

In recent years, many IoT devices including wearable devices have been limited, and there is a limit to a question and answer system that solves only simple questions.

In order to alleviate this inconvenience, it is necessary to analyze the utterance of the questioner in advance and grasp the intention of the questioner.

In order to realize the multimedia technology in the IoT environment, MPEG is making a standard in MPEG (Moving Picture Experts Group) IoMT (Internet of Media Things) group, and it is trying to include a Q & A user interface there.

For this purpose, research has been conducted on techniques for analyzing the user's utterance contents and performing processing according to utterance contents in an appropriate IoT device.

The present invention has been made to solve the conventional problems, and provides a query response system and method in an MPEG IoMT environment in which queries can be performed on various types of questions and commands that are input through various devices in an IoT environment do.

The objects of the present invention are not limited to the above-mentioned objects, and other objects not mentioned can be clearly understood by those skilled in the art from the following description.

According to an aspect of the present invention, there is provided a query response system in an MPEG IoMT environment including an IoT terminal for receiving and transmitting speech information, and receiving and providing query response result information; And performing speech analysis on the speech information transmitted from the IoT terminal according to a Moving Picture Experts Group (MPEG) IoMT (Internet of Media Things) data format and performing a query response with a query response server using the speech analysis information And providing the query response result information to the IoT terminal.

The data format of the MPEG IoMT preferably includes information on the user question type and information on which language the user's question is expressed.

The information on the user question type preferably includes information indicating a subject of the question, information indicating the focus of the question, and information indicating the meaning or purpose of the question.

Further, the focus information of the question is classified into a classification system such as " When, Where, What, Who, Why, How ", and the meaning and purpose information of the question are classified into a command request, a lexical request, Request and method requests, and so on.

Meanwhile, the data format of the MPEG IoMT may include question domain information represented by a string.

According to another aspect of the present invention, there is provided an IoT terminal comprising: an input unit for receiving input information provided by a user; A communication unit for transmitting the input speech information to the speech analysis server and receiving query response result information from the speech analysis server; And an output unit for outputting query response result information received from the speech analysis server.

And the input unit includes a microphone for receiving the user's utterance information.

Here, the input unit may include a query interface unit for outputting a user interface for receiving modal information in the form of text on a screen.

In addition, the input unit may include a camera for obtaining modal information in an image form.

The output unit may further include a screen output unit for outputting inquiry result information to a screen.

The output unit may further include an audio output unit for outputting the query response result information by voice.

Meanwhile, the utterance analysis server may include: a communication unit for performing data communication with the IoT terminal and the query response server; A speech recognition unit for recognizing speech of the speech information provided from the IoT terminal; An utterance analyzing unit for performing utterance analysis on the speech recognized speech information according to a Moving Picture Experts Group (MPEG) IoMT (Internet of Media Things) data format; And a query response call part for querying the query response server using the information analyzed by the data format of the MPEG IoMT.

The speech analysis server may further include a speech synthesis unit for converting the text-based query response result information into speech.

According to an embodiment of the present invention, it is determined whether the analyzed speech information is information for requesting a query or information for a device control command. If the analyzed Ignition information is a device control command, the IoT And a speech information determiner for transmitting the speech information to the terminal.

Meanwhile, the speech recognition unit performs language processing such as morphological analysis, object name analysis, and syntax analysis on the speech information.

The query response server performs a query analysis using the data format of the MPEG IoMT of the information received from the utterance analysis server, and provides the query response result information, which is the result of the query analysis, to the utterance analysis server .

If there are a plurality of query response results, the query response server transmits to the utterance analysis server list information set in accordance with correct answer possibility information about the query response result.

In the query response method in the MPEG IoMT environment according to an embodiment of the present invention, the speech analysis server analyzes the speech information transmitted from the IoT terminal according to the MPEG (Moving Picture Experts Group) IoMT (Internet of Media Things) data format ; Performing a query response with the query response server using the information analyzed by the utterance analysis server; And the speech analysis server providing the query response result information to the IoT terminal.

Here, the data format of the MPEG IoMT includes information on the user question type and information on which language the user's question is expressed.

The information on the user question type includes information indicating the subject of the question, information indicating the focus of the question, and information indicating the meaning or purpose of the question.

Therefore, according to one embodiment of the present invention, the user utterance provided from the IoT terminal is analyzed, and the utterance analysis is performed according to the MPEG IoMT data format to provide a query response. Thus, even in the MPEG IoMT environment, An answer service can be provided.

1 is a diagram for explaining a configuration block of a query response system in an MPEG IoMT environment according to an embodiment of the present invention.
FIG. 2 is a block diagram illustrating the IoT terminal shown in FIG. 1 of the present invention.
FIG. 3 is a block diagram of the speech analysis server 200 shown in FIG. 1. Referring to FIG.
4 is a reference diagram for explaining a voice recognition data format for voice recognition applied to the voice processing unit shown in FIG.
5 is a reference diagram for explaining a speech recognition data format used in the speech analysis server shown in FIG.
FIG. 6 is a reference diagram for explaining an IoMT query analysis packet format used in the speech analysis server shown in FIG. 1; FIG.
FIG. 7 is a reference diagram for explaining a first example of the speech analysis in the speech analysis server shown in FIG. 1;
FIG. 8 is a reference diagram for explaining a second example of speech analysis in the speech analysis server shown in FIG. 1; FIG.
9 is a reference diagram for explaining the Qfocus classification system in the speech analysis in the utterance analysis server shown in Fig.
10 is a reference diagram for explaining the QCsemanticCS classification system in the speech analysis in the speech analysis server shown in FIG.
11 is a reference diagram for explaining a speech synthesis data format used in the speech analysis server shown in Fig.
12 illustrates a configuration block for token utilization in the utterance analysis server shown in FIG. 1;
13 is a flowchart for explaining a query response method in an MPEG IoMT environment according to an embodiment of the present invention.

BRIEF DESCRIPTION OF THE DRAWINGS The advantages and features of the present invention and the manner of achieving them will become apparent with reference to the embodiments described in detail below with reference to the accompanying drawings. The present invention may, however, be embodied in many different forms and should not be construed as being limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art. Is provided to fully convey the scope of the invention to those skilled in the art, and the invention is only defined by the scope of the claims. It is to be understood that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. In the present specification, the singular form includes plural forms unless otherwise specified in the specification. As used herein, the terms " comprises, " and / or " comprising " refer to the presence or absence of one or more other components, steps, operations, and / Or additions.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. 1 illustrates a block diagram of a query response system in an environment of MPEG (Moving Picture Experts Group) (hereinafter referred to as "MPEG") IoMT (hereinafter referred to as "IoMT") environment according to an embodiment of the present invention. Fig.

1, the query response system in the MPEG IoMT environment according to an embodiment of the present invention includes an Internet of Things (IoMT) terminal 100, a speech analysis server 200, And a query response server 300.

The IoT terminal 100 receives the speech information provided from the user and transfers the speech information to the speech analysis server, and provides the user with the query response result information received from the speech analysis server.

Here, the IoT terminal 100 may include various sensors and control devices as all devices used in the IoT environment including a wearable device.

Meanwhile, the IoT terminal 100 may provide the information and sensing information of the device together with the utterance information including the query information of the user to the utterance analysis server 200 together with the utterance information.

FIG. 2 is a block diagram illustrating the IoT terminal shown in FIG. 1 of the present invention.

2, the IoT terminal 100 includes an input unit 110, a communication unit 120, and an output unit 130. As shown in FIG.

The input unit 110 receives the user's utterance information. The input unit 110 in the present embodiment is preferably a microphone that receives the speech information. However, the input unit 1000 may further include at least one or more of a query interface for outputting a user interface for inputting text-type utterance information and a camera for acquiring image-type utterance information.

The communication unit 120 transmits the input speech information to the speech analysis server 200 and receives query response result information from the speech analysis server 200. Here, the information exchanged by the communication unit 120 may include data such as voice, text, and images, a device control command as a result of the utterance analysis, a question sentence questioned by the user, and a correct answer candidate list, which is inquiry response result information.

The output unit 130 also outputs the query response result information provided from the utterance analysis server 200. The output unit 130 in this embodiment includes a screen output unit 130 for outputting question and answer result information to a screen through a user interface and at least one or more audio output units 130 for outputting question and answer result information as a voice .

The speech analysis server 200 performs speech analysis on the speech information provided from the IOT terminal 100 according to the MPEG (Moving Picture Experts Group) data format of IoMT (Internet of Media Things) Performs a query response with the response server 300 and provides the query response result information to the IoT terminal 100.

FIG. 3 is a block diagram of the speech analysis server 200 shown in FIG. 1. Referring to FIG. 3, the utterance analysis server 200 includes a first communication unit 210, a voice processing unit 220, an utterance analysis unit 230, a utterance information determination unit 240, and a query response call unit 250, And a second communication unit 270.

The first communication unit 210 performs communication with the IoT terminal 100.

The voice processing unit 220 recognizes the voice of the utterance information provided from the IoT terminal 100.

4 is a reference diagram for explaining a voice recognition data format for voice recognition applied to the voice processing unit shown in FIG.

To this end, the voice processing unit 220 includes a "Speech Recognition Type" field (voice recognition type) field for providing a summary description of voice recognition, as shown in FIG. 4 and the following Table 1, and a speech recognition data format composed of " speechText " (voice text) field. At this time, the utterance information can be subjected to general language processing such as morphological analysis, object name analysis, and syntax analysis.

Name Definition SpeechRecognitionType
(Voice recognition type) Provides an abstract description of speech recognition, which is done in the media analyzer. speechText
(Voice text) Describes the resulting text of speech recognition.

5 is a reference diagram for explaining the results of the speech analysis shown in FIG.

For example, when the analyzed data is a speech recognition result, and this is the text "Please turn to Channel 7." outputted from the voice of the user, the voice processing unit 220 outputs " SpeechRecognitionType "field contains" xai: type "and the" speechText "field contains" Please turn to Channel 7 ".

6 and Table 2, the speech analysis unit 230 performs speech analysis on speech-recognized speech information according to the MPEG (Moving Picture Experts Group) IoMT (Internet of Media Things) data format .

The speech analysis and the query analysis are made of "QuestionAnalysisType" (query analysis type), and it consists of two elements in the form of extended basic data analysis type used in MPEG IoMT.

One is the analyzed "anlyzedQuestion" (question element), the "UserQuestionType" (user question type), and the other is the language element, which analyzes the user's question expressed in what "language" (language). In other words, these two elements represent information about the analyzed question.

"UserQuestionType" (User Question Type) consists of three elements and one attribute.

The first element represents "Qtopic" (the subject of the question) and is expressed in the form of a string, and the second element represents "Qfocus" (the focus of the question).

Here, " Qfocus " (focus of the question) is classified and expressed in advance as a CQfocus classification system as shown in FIG. 9 and [Table 3].

The third element is the Q domain, which is the attribute of the question or the purpose of the question. Finally, it is possible to express the field of the question as a string. When the user's question is analyzed, the analysis result is divided into the subject, the focus, the meaning and the field of the question, and the expression format is transmitted to the appropriate module of the server or the terminal to perform the required operation.

Name Definition QuestionAnalysisType Provides an abstract description of the question analysis, which can be done in a processing unit. anlyzedQuestion Describes the result of the question analysis. language Indicates the language of the input question.
NOTE If present, the language element shall take precedence over other language indications present within the input question. UserQuestionType Provides abstracts of User Question description. Describes the user's utterance that the output speech recognition process. User Question is sent to QA server. If it is a control command, it is sent to the actuator. Qtopic Describes topic of the question. Question topic is the object or event that the question is about. Ex. Qtopic is King Lear in "Who is the King of Lear?" Qfocus Describes the focus of the question, which is one of 5W1H. The type of the focus shall be described using the mpeg7: termReferenceType defined in 7.6 of ISO / IEC 15938-5: 2003. A classification scheme that may be used for this purpose is the QfocusCS defined in this document (A.4.3) .Ex. What, where, who, what policy. qCsemantic Describes the question classification based on the meaning / purpose of the question. The type of the question classification is described using the mpeg7: termReferenceType defined in 7.6 of ISO / IEC 15938-5: 2003. A classification scheme that may be used for this purpose is the QCsemanticCS defined in this document (A.4.4). Ex. What does MPEG stand for? (Request for terminology). Could you please turn on the TV? (Request for command) qdomain Describes the domain of the question such as "science", "weather", "history". Ex. Who is the third king of Yi dynasty in Korea? (qdomain: history)

Here, the CQfocus classification scheme shows that the user's question corresponds to one of 5W1H, as shown in [Table 3] below. And we can express the question "When, where, what, who, why, how" in binary.

Binary representation Term ID of QfocusCS 0000 What_question 0001 Where_question 0010 When_question 0011 Who_question 0100 Why_question 0101 How_question 0110 to 1111 Reserved

In addition, the "QCsemanticCS" classification system can be expressed in binary with respect to the questions "command request, lexical request, semantic request, information request, method request" as shown in [Table 4] below.

Binary representation Term ID of QCsemanticCS 0000 Request_for_command 0001 Request_for_terminology 0010 Request for meaning 0011 Request for information 0100 Request for method 0101-1111 Reserved

For example, as shown in FIG. 7, if the query analysis result of the user query "Who is the author of King Lear?" Is "analyzedQuestion" and the language "en-us" The domain of the query is "Literature", the subject of the query is "King Lear", the focus of the query is "Who", and the purpose of the question is "Request_for_inforamtion".

The first question, "Who is the author of King Lear", was analyzed first that the language is English, the subject of the question is Lear King, the focus is "who" and the meaning / purpose of the question is " Request ", and it is shown that the analysis result is properly included in the format.

Referring to the second example, as shown in FIG. 8, if the query analysis result of the user query "How do you make Kimchi?" Is "analyzedQuestion" and the language "en-us" Is the domain of "Cooking", the subject of the query is "Kimchi", the focus of the question is "How", and the purpose of the question is "Request_for_method".

An example of the second question is "How do you make Kimchi?" This question is also analyzed in English. The field of the question is "cooking", the subject of the question is "kimchi", the focus of the question is "how" and the purpose of the question is analyzed as "information request" It is embedded and shared between modules.

The speech information determination unit 240 determines whether the analyzed speech information is information for requesting a query or information for a device control command. If the analyzed speech information is a device control command, the speech information determination unit 240 transmits the speech information to the corresponding IoT terminal 100 to perform the device control command.

When the analyzed speech information is query information, the query response call unit 250 transmits the information, which is ignited and analyzed in the data format of the MPEG IoMT, to the query response server 300 using the second communication unit 260 . Here, the second communication unit 260 is responsible for communicating with the query response server 300.

On the other hand, the voice synthesizer 270 transmits the inquiry result information to the question answering server 300 to the IoT terminal 100. At this time, according to the query response result information received by the query response server 300, the result of performing a query response of a text form is converted into speech using a speech synthesis data format as shown in FIG. 11, 210 to the IoT terminal 100.

As shown in Table 5, the speech synthesis data format includes a Speech SynthesisType field that provides an abstract description of speech synthesis that can be performed by the speech synthesis unit, a TextInput field that describes a text input to be synthesized in the course of speech synthesis An OutputSpeechFeature field indicating a voice output characteristic such as a gender, a tone and a voice speed reflected on the voice output at the time of voice output, and a Language field indicating a language of the voice inputted.

Name Definition SpeechSynthesisType Provides an abstract description of speech synthesis, which can be done in a processing unit. TextInput Describes text input to be synthesized by the process of speech synthesis. OutputSpeechFeature Output speech features such as gender, tones and voice speed to be reflected in speech output. Language Indicates the language of the input speech.NOTE If present, the Language element shall take precedence over other language indications.

When the query response server 300 requests a query using the data of MPEG IoMT, the query analysis server 300 analyzes the query using the query analysis information included in the data of the MPEG IoMT, And transmits the query response result information to the utterance analysis server 200.

Therefore, according to one embodiment of the present invention, the user utterance provided from the IoT terminal is analyzed, and the utterance analysis is performed according to the MPEG IoMT data format to provide a query response. Thus, even in the MPEG IoMT environment, An answer service can be provided.

The speech analysis server 200 according to an embodiment of the present invention compares the location information of the terminal transmitted from the IoT terminal 100 with the location information of the terminal (POI) stored in the database, And a location information searching unit (not shown) for recognizing the location of the user.

12 is a block diagram illustrating a configuration block for using a token in the utterance analysis server shown in FIG.

12, the speech analysis server 200 according to an embodiment of the present invention includes an API processing unit 281 for speech recognition, an API processing unit 282 for speech synthesis, and an API for query analysis, which are used in MPEG IoMT, And a processing unit 283.

The API processing unit 281 for speech recognition uses the API packet format in which the class of the IoMT speech recognizer that extends the MAnalyzer class is used, as shown in [Table 6].

Nested Classes Modifier and Type Method and Description Constructor Constructor and Description MSpeechRecognizer () Default constructor MSpeechRecognizer (String id) MSpeechRecognizer (String id, String ipAddress, Integer port) Fields Modifier and Type Field and Description Methods Modifier and Type Method and Description AnalyzedDataType GetSpeechText () This function returns a class (ie Java or C ++) or a structure (ie, C), which includes a returning type and an extracting text. AnalyzedDataType GetSpeechText (tid) This method returns a class (i.e., Java or C ++) or a structure (i.e., C), which includes a returning type and an extracted speech text. Float GetSpeechText _Cost (int tokenType, String tokenName)) This function returns the amount of tokens to use GetSpeechText (). If tokenType is 0, it means "Crypto Currency", if tokenType is 1, it means "Legal Tender". The token name is described in string (eg, term ID or binary representation) from TokenCS specified in A.5. GetSpeechText _Cost (0, "BTC") or GetSpeechText_Cost (0, "00000001")
Ex) GetSpeechText _Cost (1, "USD") or GetSpeechText _Cost (1, "10010100")

The API processing unit 282 for speech synthesis uses the API packet format in which the class of the IoMT speech synthesizer, which extends the MAnalyzer class, is used, as shown in [Table 7].

Nested Classes Modifier and Type Method and Description Constructor Constructor and Description MSpeechSynthesizer () Default constructor MSpeechSynthesizer (String id) MSpeechSynthesizer (String id, String ipAddress, Integer port) Fields Modifier and Type Field and Description Methods Modifier and Type Method and Description String GetSpeechSynthesisURI () This function returns a URI of a synthesized speech. AnalyzedDataType GetSpeechSynthesisURI (tid) This method returns a class (i.e., Java or C ++) or a structure (i.e., C), which includes a returning type. Float GetSpeechSynthesisURI_Cost (int tokenType, String tokenName)) This function returns the amount of tokens to use GetSpeechSynthesisURI (). If tokenType is 0, it means "Crypto Currency", if tokenType is 1, it means "Legal Tender". The token name is described in string (eg, term ID or binary representation) from TokenCS specified in A.5. GetSpeechSynthesisURI _Cost (0, "BTC") or GetSpeechSynthesisURI_Cost (0, "00000001")
Ex) GetSpeechSynthesisURI _Cost (1, "USD") or GetSpeechSynthesisURI _Cost (1, "10010100")

In addition, the API processing unit 283 for query analysis uses the API packet format in which the class of the IoMT query analyzer, which extends the MAnalyzer class, is used, as shown in [Table 6].

Nested Classes Modifier and Type Method and Description Constructor Constructor and Description MQuestionAnalyzer () Default constructor MQuestionAnalyzer (String id) MQuestionAnalyzer (String id, String ipAddress, Integer port) Fields Modifier and Type Field and Description Methods Modifier and Type Method and Description AnalyzedDataType GetUserQuestion () This function returns a class (ie Java or C ++) or a structure (ie, C), which includes a returning type and user question. AnalyzedDataType GetUserQuestion (tid) This method returns a class (i.e., Java or C ++) or a structure (i.e., C), which includes a returning type and user question. Float GetUserQuestion_Cost (int tokenType, String tokenName)) This function returns the amount of tokens to use GetUserQuestion (). If tokenType is 0, it means "Crypto Currency", if tokenType is 1, it means "Legal Tender". The token name is described in string (eg, term ID or binary representation) from TokenCS specified in A.5. GetUserQuestion_Cost (0, "BTC") or GetUserQuestion_Cost (0, "00000001")
Ex) GetUserQuestion_Cost (1, "USD") or UserQuestion_Cost (1, "10010100")

Accordingly, the speech analysis server has an effect of providing a transaction service whenever providing services such as query analysis, speech recognition, and speech synthesis in the MPEG IoMT environment.

Hereinafter, a query response method in the MPEG IoMT environment according to an embodiment of the present invention will be described with reference to FIG.

13 relates to a query response processing method in an MPEG IoMT environment, and is preferably performed by a speech analysis server.

First, the speech analysis server 200 receives the speech information transmitted from the IoT terminal 100 (S100).

Then, the utterance analysis server 200 performs utterance analysis on the input utterance information according to the Moving Picture Experts Group (I / O) data format of IoMT (S200). Here, the data format of the MPEG IoMT includes information on the user question type and information on which language the user's question is expressed.

The information on the user question type includes information indicating the subject of the question, information indicating the focus of the question, and information indicating the meaning or purpose of the question.

Further, the focus information of the question is classified into a classification system such as " When, Where, What, Who, Why, How ", and the meaning and purpose information of the question are classified into a command request, a lexical request, Request and method requests, and so on.

Meanwhile, the data format of the MPEG IoMT may include question domain information represented by a string.

At this time, the utterance analysis server 200 determines whether it is the result of the utterance analysis on the query as a result of the analyzed utterance analysis (S300).

In step S300, if the query is for a query (YES), the utterance analysis server 200 performs a query response with the query response server using the information analyzed using the speech (S400).

Thereafter, the utterance analysis server 200 provides the query response result information to the IoT terminal (S500).

Meanwhile, if it is determined in step S300 that the device control is ignited (YES), the contents of the speech analysis are provided to the IoT terminal 100 (S600).

Therefore, according to one embodiment of the present invention, the user utterance provided from the IoT terminal is analyzed, and the utterance analysis is performed according to the MPEG IoMT data format to provide a query response. Thus, even in the MPEG IoMT environment, An answer service can be provided.

While the present invention has been described in detail with reference to the accompanying drawings, it is to be understood that the invention is not limited to the above-described embodiments. Those skilled in the art will appreciate that various modifications, Of course, this is possible. Accordingly, the scope of protection of the present invention should not be limited to the above-described embodiments, but should be determined by the description of the following claims.

100: IoT terminal 200: Ignition analysis server
210: first communication unit 220:
230: utterance analysis unit 240: utterance information determination unit
250: query response calling unit 260: second communication unit
300: Query-response server

Claims (20)

An IoT terminal for receiving and transmitting speech information and receiving and providing query response result information; And
Performs speech analysis on the speech information transmitted from the IoT terminal according to the MPEG (Moving Picture Experts Group) IoMT (Internet of Media Things) data format, and performs a query response with a query response server And providing a query response result information to the IoT terminal.
The method according to claim 1,
The data format of the MPEG IoMT includes:
A query response system in an MPEG IoMT environment that includes information about a user question type and a user's query in which language it is expressed.
3. The method of claim 2,
The information about the user question type may include,
The information indicating the subject of the question, the information indicating the focus of the question, and the information indicating the meaning or purpose of the question, in the MPEG IoMT environment.
3. The method of claim 2,
The focus information of the above-
A query response system in the MPEG IoMT environment, which is classified as a classification system such as "when, where, what, who, why, how".
3. The method of claim 2,
The meaning and the purpose information of the above-
Wherein the information is classified into a classification system such as a command request, a lexical request, a semantic request, an information request, and a method request.
3. The method of claim 2,
The data format of the MPEG IoMT includes:
A query response system in an MPEG IoMT environment, comprising query domain information expressed in a string.
The method according to claim 1,
The IoT terminal,
An input unit for inputting the speech information provided by the user;
A communication unit for transmitting the input speech information to the speech analysis server and receiving query response result information from the speech analysis server; And
And an output unit for outputting query response result information received from the utterance analysis server in an MPEG IoMT environment.
8. The method of claim 7,
The input unit
And a query interface for outputting a user interface for receiving modal information of a text form on a screen.
8. The method of claim 7,
The input unit
And a camera for obtaining modal information in the form of an image.
8. The method of claim 7,
The input unit
And a microphone for receiving the user's utterance information.
8. The method of claim 7,
The output unit includes:
And a screen output unit for outputting the query response result information to a screen.
8. The method of claim 7,
The output unit includes:
And an audio output unit for outputting the question and answer result information by voice.
The method according to claim 1,
Wherein the speech analysis server comprises:
A communication unit for performing data communication with the IoT terminal and the query response server;
A speech recognition unit for recognizing speech of the speech information provided from the IoT terminal;
An utterance analyzing unit for performing utterance analysis on the speech recognized speech information according to a Moving Picture Experts Group (MPEG) IoMT (Internet of Media Things) data format;
And a query response call unit for querying the question answering server using the information analyzed by the data format of the MPEG IoMT.
14. The method of claim 13,
Wherein the speech analysis server comprises:
And a speech synthesizer for converting the text-based query response result information into speech.
14. The method of claim 13,
Wherein the controller determines whether the analyzed speech information is information for a request for inquiry or information for a device control command, and performs a device control command if the analyzed device is a device control command, And a judging unit for judging whether or not the MPEG IoMT environment is valid.
The method according to claim 1,
Wherein the query response server comprises:
Performing a query analysis using the data format of the MPEG IoMT of the information received from the utterance analysis server and providing the query response result information as a result of the query analysis to the utterance analysis server,
And transmits the list information set in accordance with the correct answer possibility information on the query response result to the utterance analysis server when a plurality of query response results exist.
A method of processing a query response in an MPEG IoMT environment,
Performing a speech analysis on the speech information transmitted from the IoT terminal according to a Moving Picture Experts Group (IoMT) Internet of Media Things (IoMT) data format;
Performing a query response with the query response server using the information analyzed by the utterance analysis server; And
And providing the query response result information to the IoT terminal by the speech analysis server.
18. The method of claim 17,
The data format of the MPEG IoMT includes:
Information about user question types
A method of query response in an MPEG IoMT environment, the method comprising:
19. The method of claim 18,
The information about the user question type may include,
The information indicating the subject of the question, the information indicating the focus of the question, and the information indicating the meaning or purpose of the question, in the MPEG IoMT environment.
20. The method of claim 19,
The focus information of the question is classified into a classification system such as " when, where, what, who, why,
Wherein the semantic and object information of the question is classified into a classification system such as a command request, a lexical request, a semantic request, an information request, and a method request, in the MPEG IoMT environment.

Images