CN112866487A - Method for judging call state of analog telephone set by using voice recognition technology - Google Patents

Method for judging call state of analog telephone set by using voice recognition technology Download PDF

Info

Publication number
CN112866487A
CN112866487A CN202110212947.9A CN202110212947A CN112866487A CN 112866487 A CN112866487 A CN 112866487A CN 202110212947 A CN202110212947 A CN 202110212947A CN 112866487 A CN112866487 A CN 112866487A
Authority
CN
China
Prior art keywords
module
voice
call
calling
voice recognition
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202110212947.9A
Other languages
Chinese (zh)
Inventor
雷俊智
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shanghai Mique Technology Co ltd
Original Assignee
Shanghai Mique Technology Co ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Shanghai Mique Technology Co ltd filed Critical Shanghai Mique Technology Co ltd
Priority to CN202110212947.9A priority Critical patent/CN112866487A/en
Publication of CN112866487A publication Critical patent/CN112866487A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04MTELEPHONIC COMMUNICATION
    • H04M3/00Automatic or semi-automatic exchanges
    • H04M3/22Arrangements for supervision, monitoring or testing
    • H04M3/2281Call monitoring, e.g. for law enforcement purposes; Call tracing; Detection or prevention of malicious calls

Abstract

The system is oriented to a traditional analog telephone system, and comprises a call control module, a voice processing module, a call state judgment module, a voice recognition module and a corpus module; the method comprises the steps of combining a calling process, monitoring and continuously acquiring a called side voice stream after calling is initiated by controlling interactive cooperation among modules, submitting the called side voice stream to a voice recognition module for processing by a calling state judgment module to obtain continuous text information, and judging the current calling state of a called side by combining called multi-scene corpus information in a corpus. The invention comprises a call control module, a voice processing module, a call state judging module, a voice recognition module, a corpus module and a main service control flow.

Description

Method for judging call state of analog telephone set by using voice recognition technology
Technical Field
The invention relates to the technical field of internet information and artificial intelligence, in particular to a method for judging the call state of an analog telephone by utilizing a voice recognition technology.
Background
Due to the technical system limitation and the influence of switch differentiation, the conventional analog telephone (conventional fixed telephone) is difficult to normally acquire or completely acquire the conventional call state information such as ringing, answering, no response, unavailable connection, empty number of the called party and the like in the calling process, so that the direct influence is caused on the development of various voice value-added services (such as voice artificial intelligence) depending on the calling state of the conventional analog telephone. For example, in the voice artificial intelligence service based on the traditional analog telephone, because whether the called party answers or not cannot be accurately acquired, the calling side or the platform side cannot judge whether to start the playback or not, and the like, which directly affects the processing of some key service process links.
In recent years, speech recognition technology has become mature and widely used, and the accuracy of speech recognition has reached the commercial level. The voice recognition technology can be utilized to judge the calling state of the called terminal by recognizing the voice of the called terminal and translating the voice into text information, and combining with the conventional conversation behavior and application scenes (such as ' beep ', ' no ', the number called ', and the user can sound ' feed ' after the called is answered, etc.) to judge the calling state of the called terminal and feed the calling state to the calling side or the platform side, thereby facilitating the accurate processing of the key business process related to the voice value-added service based on the traditional analog telephone side.
Disclosure of Invention
The invention provides a method for judging the calling state of an analog telephone by utilizing a voice recognition technology. The system is oriented to a traditional analog telephone system, and comprises a call control module, a voice processing module, a call state judgment module, a voice recognition module and a corpus module; the method comprises the steps of combining a calling process, monitoring and continuously acquiring a called side voice stream after calling is initiated by controlling interactive cooperation among modules, submitting the called side voice stream to a voice recognition module for processing by a calling state judgment module to obtain continuous text information, and judging the current calling state of a called side by combining called multi-scene corpus information in a corpus. The invention comprises a call control module, a voice processing module, a call state judging module, a voice recognition module, a corpus module and a main service control flow.
1. And the call control module is responsible for basic analog telephone call processing processes such as initiating a telephone call, continuing the call, hanging up the call and the like based on the traditional analog telephone.
2. And the voice processing module is responsible for monitoring the audio stream information of the voice channel of the called side. And according to the requirement of the call state judging module, the voice stream information is submitted in a continuous fragmentation/segmentation mode.
3. And the calling state judgment module is responsible for acquiring the voice stream of the called side, continuously submitting the audio stream information according to the requirements of the voice recognition module, acquiring text recognition result information fed back by the voice recognition module, performing integrity processing on the text information and the like to obtain text information corresponding to the original voice stream of the called side, calling the language database module, matching and recognizing the current text, and finally obtaining the state information of the current called call.
4. And the voice recognition module is responsible for receiving the continuous voice stream transmitted by the calling state judgment module, and can locally or call the external voice recognition capability to recognize the voice stream, convert the voice stream into text information and continuously feed the text information back to the calling state judgment module.
5. And the corpus library module is used for constructing corresponding text expected information under various calling states by combining the conventional call behaviors and application scenes. For example, the text "beep", "don't care", the number dialed by you does not exist "," don't care ", the number dialed by you is blank", etc. corresponding to the abnormal state of the called number, the text "don't care", the number dialed by you is in the call state, and the called answering state corresponds to the text "feed", "hello", etc.
6. The main service control process comprises the following steps: 1) the call control module initiates a call aiming at the called number; 2) when calling, the voice processing module starts the monitoring of the voice channel of the called side and the acquisition of the voice stream of the called side; 3) the voice recognition module continuously sends the acquired voice stream of the called side to the calling state judgment module according to the requirement agreed by the calling state judgment module; 4) the calling state judgment module continuously submits the voice stream information of the called side to the voice recognition module according to the requirements of the voice recognition module; 5) the voice recognition module receives the continuous voice stream transmitted by the calling state judgment module, can locally or call external voice recognition capability to recognize the voice stream, converts the voice stream into text information and feeds the text information back to the calling state judgment module continuously; 6) after obtaining the text recognition result information fed back by the voice recognition module, the call state judgment module performs integrity processing and the like on the text information to obtain the text information corresponding to the original voice stream of the called side, and calls the language database module to match and recognize the current text to finally obtain the state information of the current called call.
Drawings
Fig. 1 is a block diagram of a method for determining the call state of an analog telephone using voice recognition technology.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to fig. 1, in an embodiment of the present invention, a method for determining a call state of an analog telephone by using a voice recognition technology includes a call control module, a voice processing module, a call state determination module, a voice recognition module, a corpus module, and a main service control process.
1. The call control module (1) is responsible for basic analog telephone call processing procedures such as telephone call initiation, call connection, call hang-up and the like based on the traditional analog telephone.
2. And the voice processing module (2) is responsible for monitoring the audio stream information of the voice channel of the called side. And according to the requirement of the call state judging module, the voice stream information is submitted in a continuous fragmentation/segmentation mode.
3. And the calling state judgment module (3) is responsible for acquiring the voice stream of the called side, continuously submitting the audio stream information according to the requirements of the voice recognition module, acquiring text recognition result information fed back by the voice recognition module, performing integrity processing on the text information and the like to obtain text information corresponding to the original voice stream of the called side, calling the language database module, matching and recognizing the current text, and finally obtaining the state information of the current called call. Corpus-based text matching and recognition may utilize the following methods:
a. the precise matching method comprises the following steps: completely and accurately matching the recognized text information with the corpora in the corpus, and if the matching is correct, extracting the calling state corresponding to the corpus as the state information of the current called call;
b. the inclusion matching method is used for judging whether the text information is contained in a certain corpus of the corpus, and if the text information is contained in the certain corpus, extracting the calling state corresponding to the corpus as the state information of the current called call;
c. and (3) artificial intelligence judgment, namely purifying the recognized text by using a TFIDF model, training the purified text and corpus, judging the similarity of the training result, setting a similarity value (such as 0.7) to be over as that the corpus is successfully matched, and extracting the calling state corresponding to the corpus as the state information of the current called call.
4. And the voice recognition module (4) is responsible for receiving the continuous voice stream transmitted by the calling state judgment module, and can locally or call the external voice recognition capability to recognize the voice stream, convert the voice stream into text information and continuously feed the text information back to the calling state judgment module.
5. And the corpus library module (5) is used for constructing corresponding text expected information under various calling states by combining the normalized call behaviors and application scenes. The corpus structure is as follows (the content can be infinitely expanded according to the actual application scene):
call status Corpus
Called number does not exist Dudu, Dudu
Called number does not exist Ticker
Called number does not exist No match, no number called by you exists
Called numberCode absence The number dialed by you is a blank number
The called number is shut down If the number called by you is wrong, the machine is stopped due to arrearage
The called party is in the process of communication In the wrong place, the number dialed by you is in the process of calling, and then the number is dialed later
6. The main service control process comprises the following steps: 1) the call control module (1) initiates a call aiming at a called number; 2) when calling, the voice processing module (2) starts the monitoring of the voice channel of the called side and the acquisition of the voice stream of the called side; 3) the voice recognition module (2) continuously sends the acquired voice stream of the called side to the calling state judgment module (3) according to the requirement agreed by the calling state judgment module (3); 4) the calling state judgment module (3) continuously submits the voice stream information of the called side to the voice recognition module (4) according to the requirement of the voice recognition module (4); 5) the voice recognition module (4) receives the continuous voice stream transmitted by the calling state judgment module (3), can locally or call external voice recognition capability to recognize the voice stream, converts the voice stream into text information and continuously feeds back the text information to the calling state judgment module (3); 6) after obtaining the text recognition result information fed back by the voice recognition module, the call state judgment module (3) performs integrity processing and the like on the text information to obtain the text information corresponding to the original voice stream of the called side, and invokes the corpus module (5) to match and recognize the current text to finally obtain the state information of the current called call.

Claims (8)

1. A method for judging the calling state of an analog telephone by utilizing the voice recognition technology is characterized in that: the system is oriented to a traditional analog telephone system, and comprises a call control module, a voice processing module, a call state judgment module, a voice recognition module and a corpus module; the method comprises the steps of combining a calling process, monitoring and continuously acquiring a called side voice stream after calling is initiated by controlling interactive cooperation among modules, submitting the called side voice stream to a voice recognition module for processing by a calling state judgment module to obtain continuous text information, and judging the current calling state of a called side by combining called multi-scene corpus information in a corpus.
2. The invention comprises a call control module, a voice processing module, a call state judging module, a voice recognition module, a corpus module and a main service control flow.
3. The method of claim 1, wherein the method comprises the steps of: and the call control module is responsible for basic analog telephone call processing processes such as initiating a telephone call, continuing the call, hanging up the call and the like based on the traditional analog telephone.
4. The method of claim 1, wherein the method comprises the steps of: the voice processing module is responsible for monitoring the audio stream information of the voice channel of the called side; and according to the requirement of the call state judging module, the voice stream information is submitted in a continuous fragmentation/segmentation mode.
5. The method of claim 1, wherein the method comprises the steps of: and the calling state judgment module is responsible for acquiring the voice stream of the called side, continuously submitting the audio stream information according to the requirements of the voice recognition module, acquiring text recognition result information fed back by the voice recognition module, performing integrity processing on the text information and the like to obtain text information corresponding to the original voice stream of the called side, calling the language database module, matching and recognizing the current text, and finally obtaining the state information of the current called call.
6. The method of claim 1, wherein the method comprises the steps of: and the voice recognition module is responsible for receiving the continuous voice stream transmitted by the calling state judgment module, and can locally or call the external voice recognition capability to recognize the voice stream, convert the voice stream into text information and continuously feed the text information back to the calling state judgment module.
7. The method of claim 1, wherein the method comprises the steps of: the corpus library module is used for constructing corresponding text expected information under various calling states by combining the conventional conversation behaviors and application scenes; for example, the text "beep", "don't care", the number dialed by you does not exist "," don't care ", the number dialed by you is blank", etc. corresponding to the abnormal state of the called number, the text "don't care", the number dialed by you is in the call state, and the called answering state corresponds to the text "feed", "hello", etc.
8. The method of claim 1, wherein the method comprises the steps of: the main service control process comprises the following steps: 1) the call control module initiates a call aiming at the called number; 2) when calling, the voice processing module starts the monitoring of the voice channel of the called side and the acquisition of the voice stream of the called side; 3) the voice recognition module continuously sends the acquired voice stream of the called side to the calling state judgment module according to the requirement agreed by the calling state judgment module; 4) the calling state judgment module continuously submits the voice stream information of the called side to the voice recognition module according to the requirements of the voice recognition module; 5) the voice recognition module receives the continuous voice stream transmitted by the calling state judgment module, can locally or call external voice recognition capability to recognize the voice stream, converts the voice stream into text information and feeds the text information back to the calling state judgment module continuously; 6) after obtaining the text recognition result information fed back by the voice recognition module, the call state judgment module performs integrity processing and the like on the text information to obtain the text information corresponding to the original voice stream of the called side, and calls the language database module to match and recognize the current text to finally obtain the state information of the current called call.
CN202110212947.9A 2021-02-26 2021-02-26 Method for judging call state of analog telephone set by using voice recognition technology Pending CN112866487A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110212947.9A CN112866487A (en) 2021-02-26 2021-02-26 Method for judging call state of analog telephone set by using voice recognition technology

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110212947.9A CN112866487A (en) 2021-02-26 2021-02-26 Method for judging call state of analog telephone set by using voice recognition technology

Publications (1)

Publication Number Publication Date
CN112866487A true CN112866487A (en) 2021-05-28

Family

ID=75989970

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110212947.9A Pending CN112866487A (en) 2021-02-26 2021-02-26 Method for judging call state of analog telephone set by using voice recognition technology

Country Status (1)

Country Link
CN (1) CN112866487A (en)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105323744A (en) * 2014-06-23 2016-02-10 中兴通讯股份有限公司 Method and apparatus for call state feedback, and terminal
CN105979106A (en) * 2016-06-13 2016-09-28 北京容联易通信息技术有限公司 Ring tone recognition method and system for call center system
CN109151220A (en) * 2018-09-11 2019-01-04 中国—东盟信息港股份有限公司 A kind of communication session call failure scene analysis system
CN109697243A (en) * 2019-02-01 2019-04-30 网易(杭州)网络有限公司 Ring-back tone clustering method, device, medium and calculating equipment
CN110166637A (en) * 2018-02-12 2019-08-23 深圳市六度人和科技有限公司 A kind of spacing recognition methods and device
CN111435960A (en) * 2018-12-25 2020-07-21 马上消费金融股份有限公司 Method, system, device and computer storage medium for identifying user number state

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105323744A (en) * 2014-06-23 2016-02-10 中兴通讯股份有限公司 Method and apparatus for call state feedback, and terminal
CN105979106A (en) * 2016-06-13 2016-09-28 北京容联易通信息技术有限公司 Ring tone recognition method and system for call center system
CN110166637A (en) * 2018-02-12 2019-08-23 深圳市六度人和科技有限公司 A kind of spacing recognition methods and device
CN109151220A (en) * 2018-09-11 2019-01-04 中国—东盟信息港股份有限公司 A kind of communication session call failure scene analysis system
CN111435960A (en) * 2018-12-25 2020-07-21 马上消费金融股份有限公司 Method, system, device and computer storage medium for identifying user number state
CN109697243A (en) * 2019-02-01 2019-04-30 网易(杭州)网络有限公司 Ring-back tone clustering method, device, medium and calculating equipment

Similar Documents

Publication Publication Date Title
US7844454B2 (en) Apparatus and method for providing voice recognition for multiple speakers
US7657005B2 (en) System and method for identifying telephone callers
US11710488B2 (en) Transcription of communications using multiple speech recognition systems
CN104184872A (en) Crank call preventing mobile phone achieving method based on conversation content
US20030152199A1 (en) Dialogue device for call screening and Classification
CN111739519A (en) Dialogue management processing method, device, equipment and medium based on voice recognition
CN109348077A (en) A kind of telephone system and application method manually switched with robot
CN104618615B (en) A kind of TeleConference Bridge meeting summary method for pushing based on short message
US20210249007A1 (en) Conversation assistance device, conversation assistance method, and program
CN113779217A (en) Intelligent voice outbound service method and system based on human-computer interaction
CN105007365B (en) A kind of extension number dialing method and device
EP2913822B1 (en) Speaker recognition
CN112866487A (en) Method for judging call state of analog telephone set by using voice recognition technology
CN116340482A (en) Multi-skill customer service auxiliary product based on enterprise WeChat combined with NLP engine
CN113542509B (en) Emergency processing method, device, storage medium and equipment
CN100502536C (en) Method for realizing calling remind in grouped telecommunication
CN111901488B (en) Method for improving outbound efficiency of voice robot based on number state
US20210074296A1 (en) Transcription generation technique selection
US20220124193A1 (en) Presentation of communications
CN113641801A (en) Control method and system of voice scheduling system and electronic equipment
KR100336994B1 (en) The system and method for speech recognition potal service using multi-step speech recognition
CN111756910A (en) Incoming call disturbance-free method, system, computer equipment and storage medium
CN110534084A (en) Intelligent voice control method and system based on FreeWITCH
CN110677541B (en) System and method for answering call with assistance of artificial intelligence
CN114584656B (en) Streaming voice response method and device and voice call robot thereof

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination