Voice processing method and device, computer equipment, storage medium and program product

Abstract

The invention provides a voice processing method and device, computer equipment, a storage medium and a program product. The voice processing method comprises the following steps: receiving call voice sent by a telephone switching device; under the condition that a call ending instruction sent by the telephone switching device is received, if it is determined that the first text of the call voice is not sent to the telephone switching device, a connection request is sent to a second server, and a communication link between the first server and the second server is established; and sending the first text to the second server through the communication link. According to the method, the integrity of the text received by the second server can be ensured.

Description

Technical Field

[0001] The present application relates to the field of communications, and in particular to a speech processing method, apparatus, computer equipment, storage medium and program product. Background Art

[0002] The intelligent outbound calling system is becoming more and more popular in the market. The system can actively call customers, conduct multiple rounds of voice conversations with customers by broadcasting real-person voice messages, and collect voice conversation information. At present, the intelligent outbound calling system has a large number of dial-ups per day. If there are abnormal situations such as network fluctuations, the text of the collected call voice may be incomplete. Summary of the invention

[0003] The embodiments of the present application disclose a voice processing method, apparatus, computer equipment, storage medium and program product, which solve the technical problem in the related art that the text corresponding to the call voice obtained by the intelligent outbound call system is not complete enough under network fluctuations or other abnormal situations.

[0004] In a first aspect, the present application provides a voice processing method, which is applied to a first server, and the method includes: receiving a call voice sent by a telephone exchange device; upon receiving a call end instruction sent by the telephone exchange device, if it is determined that a first text of the call voice has not been sent to the telephone exchange device, sending a connection request to a second server and establishing a communication link between the first server and the second server; and sending the first text to the second server via the communication link.

[0005] In the voice processing method provided in the present application, it is applied to the first server, and receiving the call voice sent by the telephone exchange device indicates that the first server has received the voice spoken by the user. In the case of receiving the call end instruction sent by the telephone exchange device, it indicates that a hang-up event has occurred. In the case of a hang-up event, if the first server determines that the first text of the call voice has not been sent to the telephone exchange device, the first server determines that there is a missing text that has not been sent to the second server through the telephone exchange device. Therefore, in the case of determining that the first text of the call voice has not been sent to the telephone exchange device, a connection request is sent to the second server, and a communication link is established between the first server and the second server. Sending the first text to the second server through the communication link can ensure that the second server receives the complete text of the user's call voice.

[0006] In a second aspect, the present application provides a voice processing method, applied to a second server, comprising: establishing a communication link between the second server and the first server when receiving a call end instruction sent by a telephone exchange device and receiving a connection request sent by a first server; and receiving a first text sent by the first server through the communication link.

[0007] In the voice processing method provided in the present application, the second server is applied to determine the presence of a hang-up event when receiving a call end instruction sent by a telephone exchange device. When receiving a connection request sent by the first server, it is determined that there is text that has not been sent in time by the first server, and a communication link is established between the second server and the first server. The first text sent by the first server is received through the communication link to ensure that the text corresponding to the complete call voice is received, and a complete data set is provided for subsequent downstream tasks, which can improve the processing accuracy of downstream tasks to a certain extent.

[0008] In the third aspect, the present application also provides a voice processing device, applied to a first server, comprising: a receiving module, used to receive a call voice sent by a telephone exchange device; a communication link establishing module, used to, upon receiving a call end instruction sent by the telephone exchange device, if it is determined that the first text of the call voice has not been sent to the telephone exchange device, send a connection request to a second server and establish a communication link between the first server and the second server; a sending module, used to send the first text to the second server via the communication link.

[0009] In a fourth aspect, the present application also provides a voice processing device, applied to a second server, comprising: a first receiving module, used to establish a communication link between the second server and the first server when receiving a call end instruction sent by a telephone exchange device and a connection request sent by the first server; a second receiving module, used to receive a first text sent by the first server through the communication link.

[0010] In a fifth aspect, the present application also provides a computer device, comprising a processor and a memory, wherein the processor is configured to implement the speech processing method when executing a computer program stored in the memory.

[0011] In a sixth aspect, the present application also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the speech processing method is implemented.

[0012] In a seventh aspect, the present application also provides a computer program product comprising instructions, wherein the computer program product stores instructions that, when executed on a computer, enable the computer to implement the speech processing method. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 It is a schematic diagram of an application scenario of the speech processing method provided in an embodiment of the present application.

[0014] Figure 2 It is a flow chart of the speech processing method provided in an embodiment of the present application.

[0015] Figure 3 It is a schematic diagram of the speech processing method provided in an embodiment of the present application.

[0016] Figure 4 This is a flowchart of a speech processing method provided by another embodiment of the present application.

[0017] Figure 5 It is a schematic diagram of the interaction between the telephone exchange device, the first server and the audio processing device provided in the embodiment of the present application.

[0018] Figure 6 This is a flowchart of a speech processing method provided by another embodiment of the present application.

[0019] Figure 7 This is a flowchart of a speech processing method provided by another embodiment of the present application.

[0020] Figure 8 This is a schematic diagram of storing the first text provided in an embodiment of the present application.

[0021] Fig. 9 This is a schematic diagram of storing the first text provided by yet another embodiment of the present application.

[0022] Fig.10 It is a data processing diagram of the second server provided in an embodiment of the present application.

[0023] Fig.11 It is a structural diagram of the speech processing device provided in an embodiment of the present application.

[0024] Fig.12 It is a structural diagram of a speech processing device provided in yet another embodiment of the present application.

[0025] Fig.13 It is a schematic diagram of the structure of a computer device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0026] To facilitate understanding, some illustrations of concepts related to the embodiments of the present application are given by way of example for reference.

[0027] It should be noted that in this application, "at least one" means one or more, and "more than one" means two or more than two. "And / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone, where A and B can be singular or plural. The terms "first", "second", "third", "fourth", etc. (if any) in the specification, claims and drawings of this application are used to distinguish similar objects, rather than to describe a specific order or sequence.

[0028] In the related art, the intelligent outbound call system has a large number of dial-ups per day. In the case of network fluctuations, the collected call voice text may be incomplete. Such call voice text is usually used to process downstream tasks, such as intent recognition tasks. If the text is incomplete, it will affect the accuracy of downstream tasks.

[0029] In view of the above problems, in order to solve the technical problem that the text integrity of the call voice corresponding to the intelligent outbound call system obtained by the relevant technology is not high under network fluctuations or other abnormal situations. The embodiment of the present application provides a voice processing method, device, computer equipment, storage medium and program product, which can detect whether there is a first text that has not been sent to the telephone exchange device when a call end instruction sent by the telephone exchange device is received, so as to avoid missing text. When it is determined that the first text has not been sent to the telephone exchange device, a communication link is established between the first server and the second server, and the first text is sent to the second server through the communication link. The technical problem that the text integrity of the call voice received by the second server is not high is solved.

[0030] In order to better understand the speech processing method, apparatus, computer equipment, storage medium and program product provided in the embodiments of the present application, the application scenario of the speech processing method of the present application is first described below.

[0031] Figure 1 1 is a schematic diagram of an application scenario of the voice processing method provided in an embodiment of the present application. The application scenario includes but is not limited to: a first server 10, an audio processing device 20, a telephone exchange device 30, a second server 40, a call center 50, a call terminal 60 and a user terminal 70.

[0032] For ease of understanding Figure 1 The technical solutions included in the application scenario shown are first Figure 1 The relevant nouns involved are explained.

[0033] The first server 10 may be a Media Resource Control Protocol (MRCP) server, which is used to control the reception and transmission of real-time call voices, specifically, receive the call voices sent from the telephone exchange device 30 , and send the call voices to the audio processing device 20 .

[0034] The audio processing device 20 may be an Automatic Speech Recognition (ASR) server, which is used to recognize the call voice through speech signal processing and pattern recognition, and convert the recognized call voice into text (e.g., the first text). The call voice processed by the audio processing device 20 comes from the first server 10.

[0035] The telephone exchange device 30 may be an open source software telephone exchange (Freeswitch), which is used to receive the call voice generated by the real-time outbound call and forward the call voice to the first server 10. The telephone exchange device 30 has a built-in lua scripting language, which is the medium for the telephone exchange device 30 to communicate with the second server 40.

[0036] The second server 40 may be a server corresponding to the calling terminal 60, and may be an independent server or a server cluster formed by multiple servers. The second server 40 may have a built-in database, an intelligent outbound calling platform, and a timing scheduling platform for starting and managing the call service and managing the data related to the call service.

[0037] The call center 50 may be an Internet call center (ICC), which is used to detect the call voice generated when the call terminal 60 makes an outbound call or the user terminal 70 makes an inbound call. The outbound call refers to a phone call actively initiated by the call terminal 60, and the inbound call refers to a phone call actively initiated by the user terminal 70 and passively answered by the call terminal 60. The call center 50 forwards the detected call voice to the telephone exchange device 30, so that the telephone exchange device 30 sends a corresponding instruction to the first server 10 based on the sender of the call voice.

[0038] The calling terminal 60 may be a computer device such as a computer, a mobile phone, a notebook, a dedicated handheld device, a wearable device, etc. The calling terminal 60 may also be a robot. The calling terminal 60 may be installed with online customer service software or a program supporting multi-channel access. Based on the installed online customer software or program, telephone services such as telephone marketing services and customer service hotline services can be realized.

[0039] The user terminal 70 may be a computer, a mobile phone, a notebook, a wearable device, a handheld device or other electronic device with communication function.

[0040] like Figure 1 In the application scenario shown, if the user corresponding to the user terminal 70 has the need to consult related services, he can actively initiate a call to the call terminal 60 (for the call terminal 60, that is, "incoming call"). If the call terminal 60 has the need to promote and market related products, it can actively initiate a call to the user terminal 70 (for the call terminal 60, that is, "outgoing call"). After the incoming call or outgoing call, a communication link is established, and the call center 50 can obtain the call voice generated during the call in real time, and send the obtained call voice to the telephone exchange device 30. When the call voice is received, if the telephone exchange device 30 detects that the call voice is the call voice issued by the user terminal 70, the call voice and the voice detection instruction are sent to the first server 10. After receiving the call voice and the voice detection instruction, the first server 10 sends the call voice to the audio processing device 20. The audio processing device 20 converts the call voice into text.

[0041] After the telephone exchange device 30 determines that a round of voice conversation has ended, it sends a stop detection instruction to the first server 10. In response to the stop detection instruction, the first server 10 triggers the audio processing device 20 to feed back the text corresponding to the voice conversation of this round of conversation to the first server 10. The first server 10 sends the text corresponding to the voice conversation to the second server 40 through the telephone exchange device 30 for storage and analysis.

[0042] After the telephone exchange device 30 determines that the call is ended (the hang-up event occurs), it sends a call end instruction to the first server 10. If the first server 10 does not receive the text corresponding to the call voice sent by the audio processing device 20 before receiving the call end instruction, it is determined that there is text that has not been sent to the telephone exchange device 30. The first server 10 establishes a communication link between the first server 10 and the second server 40 by calling the outbound call platform service in the second server 40, and sends the unsent text to the second server 40, so that the second server 40 receives the complete text corresponding to the call voice.

[0043] In the case of a hang-up event, after the second server 40 receives the text sent by the first server 10, it provides a complete data basis for executing downstream tasks. The downstream tasks may be intent recognition, sentiment analysis, customer behavior analysis, etc.

[0044] Based on the above application scenario architecture, the embodiments of the present application provide a speech processing method, apparatus, computer equipment, storage medium and program product. Figure 2 is a flow chart of a speech processing method provided in an embodiment of the present application, Figure 2 The method in can be Figure 1The first server 10 in the flowchart is executed. According to different requirements, the order of the steps in the flowchart can be changed, and some steps can be omitted.

[0045] Step S201, receiving a conversation voice sent by a telephone exchange device.

[0046] In some embodiments of the present application, the call voice is the voice of the target user during the call, which comes from the user terminal corresponding to the target user. There may be multiple rounds of call voice during the call. For example, if multiple rounds of conversations between the user and the calling terminal are detected multiple times, the call voices corresponding to the multiple rounds of users will be detected accordingly. The first server receives the call voice of the telephone exchange device, indicating that the user is detected to be speaking. Among them, the call voice carries a call identifier, which can be a buried point set by the second server when triggering the calling terminal to perform an outbound call operation. The buried point is used to mark the call voices belonging to the same call and to distinguish different calls.

[0047] In addition, in addition to receiving the call voice from the telephone exchange device, the first server also receives a voice detection instruction sent by the telephone exchange device. The voice detection instruction is used to trigger the first server to send the call voice to the audio processing device for voice recognition. Among them, at least one voice recognition algorithm can be pre-set in the audio processing device, and this application is not limited to this.

[0048] Step S202, when receiving the call end instruction sent by the telephone exchange device, if it is determined that the first text of the call voice is not sent to the telephone exchange device, a connection request is sent to the second server, and a communication link between the first server and the second server is established.

[0049] In some embodiments of the present application, the call end instruction indicates that a hang-up event (call end) occurs. The call end instruction carries a call identifier, and when the first server receives the call end instruction, it determines that the communication link corresponding to the call identifier is disconnected. Figure 1 As shown, when the first server 10 receives the call end instruction, it indicates that the communication link between the telephone exchange device 30 and the first server 10 is disconnected.

[0050] In some embodiments of the present application, Figure 1As shown, if the call event is normal, before the communication link between the telephone exchange device 30 and the first server 10 is disconnected, the first server 10 will receive the first text sent by the audio processing device 20, and send the first text to the second server 40 through the telephone exchange device 30. If the call event is abnormal, after the communication link between the telephone exchange device 30 and the first server 10 is disconnected, the first server 10 cannot send the first text to the second server 40 through the telephone exchange device 30, resulting in the first server 10 determining that the first text of the call voice is not sent to the telephone exchange device 30.

[0051] In some embodiments of the present application, when it is determined that the first text exists, in order to send the first text to the second server and avoid the situation where the text stored in the second server is incomplete and the last sentence text is omitted, a connection request is sent to the second server, and a communication link between the first server and the second server is established.

[0052] Step S203: sending the first text to the second server via the communication link.

[0053] In some embodiments of the present application, the first server can call the outbound call platform service through a connection request, and the outbound call platform service can be a corresponding service of the intelligent outbound call platform of the second server. Generally, the telephone exchange device supports multiple protocols, and using the telephone exchange device as a transfer device can ensure the protocol compatibility between different systems. In the case where the communication link between the first server and the telephone exchange device is disconnected, the first server can establish a communication link with the second server by calling the outbound call platform and send the first text to the second server.

[0054] Figure 3 Schematic diagram of the speech processing method provided in the embodiment of the present application. Figure 3 As shown, a call is used as an example for explanation. When the first server 10 receives the call end instruction, it is unable to transfer data to the second server 40 through the telephone exchange device 30. The first server 10 sends the first text sent by the audio processing device 20 to the second server 40 by calling the outbound call platform service of the second server 40.

[0055] In an embodiment of the present application, receiving a call voice sent by a telephone exchange device indicates that the first server has received the voice of the user. In the case of receiving a call end instruction sent by the telephone exchange device, it indicates that a hang-up event has occurred. In the case of a hang-up event, if the first server determines that the first text of the call voice has not been sent to the telephone exchange device, the first server determines that there is a missing text that has not been sent to the second server through the telephone exchange device. Therefore, in the case of determining that the first text of the call voice has not been sent to the telephone exchange device, a connection request is sent to the second server, and a communication link is established between the first server and the second server. Sending the first text to the second server through the communication link can ensure that the second server receives the complete text of the user's call voice.

[0056] Figure 4 This is a flowchart of a speech processing method provided by another embodiment of the present application. Figure 5 1 is a schematic diagram of the interaction between the telephone exchange device, the first server and the audio processing device provided by the embodiment of the present application. When a voice detection instruction and a call end instruction are received, it is determined whether the call event is abnormal. Figure 5 To explain, Figure 4 As shown, the following steps are included.

[0057] Step S401: The first server determines whether it has received a stop detection instruction before receiving a call end instruction.

[0058] In some embodiments of the present application, the voice detection instruction indicates the beginning of a round of call voice, the stop detection instruction indicates the end of a round of call voice, and the voice detection instruction is used in combination with the stop detection instruction. The call end instruction indicates the end of the call. In the case of the call end, in order to avoid the situation of text omission, the first server can detect whether the stop detection instruction is received before receiving the call end instruction. The stop detection instruction is used to trigger the first server to request the audio processing device to send the text of the call voice, such as the first text.

[0059] In order to facilitate understanding of the interactive relationship between the voice detection command, the stop detection command and the call end command, as well as the normal call events, the first type of abnormal call events and the second type of abnormal call events generated based on the interactive relationship, the following is a combination of Figure 5 To explain.

[0060] like Figure 5As shown, the situation of a normal call event is specifically as follows: after a round of call voice is established, the telephone exchange device 30 sends a voice detection instruction (vad_on) to instruct the first server 10 to start entering the voice recognition process. In response to vad_on, the first server 10 triggers the audio processing device 20 to recognize the call voice, and the audio processing device 20 recognizes the call voice. After a round of call voice ends, the telephone exchange device 30 sends a stop detection instruction (vad_off) to trigger the first server 10 to request the audio processing device 20 to send the first text. In response to vad_off, the first server 10 requests the audio processing device 20 to stop recognizing and feedback the first text. The audio processing device 20 sends the first text generated by recognizing the call voice to the first server 10. The first server 10 sends the first text to the telephone exchange device 30 for transfer. When the call ends, the telephone exchange device 30 sends a call end instruction (sip bye) to the first server 10.

[0061] like Figure 5 As shown, the first type of abnormal event is specifically as follows: after a round of call voice is established, the telephone exchange device 30 sends a voice detection instruction (vad_on) to instruct the first server 10 to start entering the voice recognition process. In response to vad_on, the first server 10 triggers the audio processing device 20 to recognize the call voice, and the audio processing device 20 recognizes the call voice. After a round of call voice ends, the telephone exchange device 30 sends a stop detection instruction (vad_off) to trigger the first server 10 to request the audio processing device 20 to feedback the first text. In response to the hang-up event (the event generated by the end of the call), the telephone exchange device 30 sends a sip bye to the first server 10 to disconnect the communication link between the telephone exchange device 30 and the first server 10. In the case of disconnecting the communication link between the telephone exchange device 30 and the first server 10, the first server 10 receives the first text sent by the audio processing device 20, but cannot send the first text to the second server through the telephone exchange device 30, resulting in the first server 10 having the first text that has not been sent to the telephone exchange device 30, and it is determined that the first type of abnormal event exists.

[0062] like Figure 5As shown, the second type of abnormal event is specifically as follows: after a round of call voice is established, the telephone exchange device 30 sends a voice detection instruction (vad_on) to instruct the first server 10 to start entering the voice recognition process. The first server 10 responds to vad_on and triggers the audio processing device 20 to recognize the call voice. In response to the hang-up event, the telephone exchange device 30 sends a sip bye to the first server 10 to disconnect the communication link between the telephone exchange device 30 and the first server 10. Since the first server 10 does not receive vad_off, it cannot trigger the audio processing device 20 to feedback the first text, resulting in the first text being omitted in the audio processing device 20, and it is determined that the first server 10 has the first text that has not been sent to the telephone exchange device 30, and it is determined that the first type of abnormal event exists. The first server determines whether it has received a stop detection instruction before receiving a call end instruction.

[0063] In some embodiments of the present application, Figure 5 As shown, if the stop detection instruction is not received before the call end instruction is received (such as Figure 5 If the first type of abnormal event is detected, the step S402 is executed. If the stop detection instruction is received before the call end instruction is received, the step S403 is executed.

[0064] Step S402: determine whether a first type of abnormal event exists.

[0065] In some embodiments of the present application, if a stop detection instruction is not received before a call end instruction is received, it means that the first server has not yet triggered the audio processing device to feedback the first text, and the first text is still stored in the audio processing device. In order to facilitate the subsequent first server to transmit the first text back to the second server, the first server can send a return instruction to the audio processing device, and the return instruction is used to instruct the audio processing device to send the first text. The first server can determine whether the first text sent by the audio processing device has been received through the call identifier. If the text sent by the audio processing device carries the call identifier, it means that the text is the first text.

[0066] Step S403: determining whether the first text sent by the audio processing device is received.

[0067] In some embodiments of the present application, if the first text sent by the audio processing device is not received, step S404 is performed. If the first text sent by the audio processing device is received, step S405 is performed.

[0068] Step S404: determine whether a second type of abnormal event exists.

[0069] In some embodiments of the present application, the first server receives a stop detection instruction before receiving a call end instruction, and the stop detection instruction is used to trigger the first server to request the audio processing device to feedback the first text. Since it takes a certain amount of time for the audio processing device to feedback data, if the first text (such as the first text sent by the audio processing device) has not been received when the call end instruction is received, Figure 5 or if, after a preset time after receiving the call end instruction, a first text fed back by the audio processing device is received, it indicates that a second type of abnormal event exists.

[0070] In addition, if the call end instruction and the first text are received at the same time, the first server cannot send the first text to the telephone exchange device. This situation also belongs to the second type of abnormal event.

[0071] Step S405: determine that the call event is normal.

[0072] In the above embodiment, it is possible to avoid the situation where the last sentence of the call (such as the first text) is omitted in the first server or the audio processing device in the case of abnormal call.

[0073] Figure 6 is a flowchart of a speech processing method provided by another embodiment of the present application. Figure 6 The method in can be Figure 1 The second server 40 in the embodiment executes. Figure 6 The process shown includes the following steps.

[0074] Step S601: upon receiving a call ending instruction sent by a telephone exchange device and a connection request sent by a first server, a communication link is established between a second server and a first server.

[0075] In some embodiments of the present application, multiple outbound call tasks may be pre-stored in the second server. A corresponding call identifier is created for each outbound call task. Based on the outbound call task, when the telephone exchange device detects the call voice corresponding to the user terminal, it sends a voice detection instruction (vad_on) to the first server. The voice detection instruction is used to trigger the first server to send the call voice to the audio processing device, and trigger the audio processing device to recognize the call voice. After determining that a section of call voice has ended, the telephone exchange device sends a stop detection instruction (vad_off) to the first server. The stop detection instruction is used to trigger the first server to request the audio processing device to feedback the text corresponding to the call voice. The second server receives the first text through the telephone exchange device, and stores and analyzes the first text.

[0076] When a telephone hang-up event occurs, the telephone exchange device sends a call end instruction (sip bye) to the first server and the second server, and the call end instruction indicates that a hang-up event has occurred. The second server receives the call end instruction sent by the telephone exchange device and determines that a hang-up event has occurred, or the second server determines that a hang-up event has occurred when detecting a hang-up operation.

[0077] In some embodiments of the present application, when the second server determines that a hang-up event occurs, it receives a connection request sent by the first server, and the connection request is used to schedule an outbound call platform service of the outbound call platform. Based on the connection request, the second server determines that there is an unsent text on the first server. A communication link is established between the second server and the first server.

[0078] Step S602: receiving a first text sent by a first server through a communication link.

[0079] In some embodiments of the present application, a first text sent by a first server is received through a communication link to ensure the integrity of the text of the call voice stored by a second server.

[0080] In the above embodiment, when there is a hang-up event, the second server receives the first text sent by the first server, which can ensure the integrity of the text of the call voice.

[0081] Figure 7 is a flow chart of a voice processing method provided by another embodiment of the present application. A timing scheduling platform (Mdes) is provided in the second server. According to a preset scheduling cycle, the second server detects in real time whether the first text is received. When the first text is received, the first text is stored in the call list. The storage method of the first text is as follows: Figure 7 As shown, the following steps are included.

[0082] Step S701: searching a call list for a subtext of a second text having a call identifier.

[0083] In some embodiments of the present application, the second server includes a database for storing data. In the database, each list has a corresponding identifier for distinguishing calls with different users. Based on the call identifier, a call list with the call identifier can be queried in the database, and the call list is used to store the second text corresponding to the target user.

[0084] The storage rule of the call list is to store the call in rounds, wherein the call speech from the time when the user starts speaking to the time when the user stops speaking is detected is one round. The second text may include texts of multiple rounds of call speech, and the subtext of the second text is the text corresponding to the last round of call speech in the call list.

[0085] In one example, based on the storage order, the call column includes the text of the first round of call speech, the text of the second round of call speech, and the text of the third round of call speech, and the text of the third round of call speech is the sub-text.

[0086] Step S702, determining whether the sub-text is the text of the call speech corresponding to the target user.

[0087] In some embodiments of the present application, the storage method of the first text in the call list can be determined by judging whether the subtext is the voice text corresponding to the target user. If the subtext is the text of the call voice corresponding to the target user, step S703 is executed. If the subtext is not the text of the call voice corresponding to the target user, step S706 is jumped to execute.

[0088] Step S703, determining that the subtext is in the first list area of ​​the call list.

[0089] In some embodiments of the present application, the storage location of the subtext in the call list is used as the first list area.

[0090] Step S704, concatenating the subtext and the first text to obtain a concatenated text.

[0091] In some embodiments of the present application, if the sub-text is the text of the call speech corresponding to the target user, the first text can be directly concatenated after the sub-text to obtain a concatenated text.

[0092] Step S705, storing the concatenated text in the first list area.

[0093] In some embodiments of the present application, the concatenated text is stored in the first list area to ensure the integrity of the text of the call list record.

[0094] Figure 8 is a schematic diagram of storing the first text provided in the embodiment of the present application. Figure 8 As shown, "XXX" represents other subtexts of the second text pre-stored in the call list. Before the first text is stored, the subtext is located in the first list area, and the first list area is the last list area of ​​the call list. After the first text is stored, the concatenated text containing the subtext and the first text is stored in the first list area.

[0095] Step S706: Create a second list area in the call list.

[0096] In some embodiments of the present application, if the target content is not the text of the call voice corresponding to the target user, it may be the voice text corresponding to the robot or the calling terminal. In order to distinguish the voice texts of different objects, a second list area is created in the call list, and the second list area is after the first list area.

[0097] Step S707: store the first text in the second list area.

[0098] Fig. 9 FIG. 1 is a schematic diagram of storing a first text provided by another embodiment of the present application. Fig. 9 As shown, "XXX" represents other subtexts of the second text pre-stored in the call list. Before the first text is stored, the subtext is located in the first list area, which is the last list area of ​​the call list. After the first text is stored, the first text is stored in the second list area, which is a newly added list area of ​​the call list and is located after the first list area.

[0099] In the above embodiment, the voice texts corresponding to the user terminal and the calling terminal are distinguished by partition storage, and the objects corresponding to the stored voice texts are clearly defined, which improves the analysis efficiency of the original call list to a certain extent.

[0100] In other embodiments of the present application, after the first text is stored in the call list, the first text and the second text are used as target texts. The second server calls a pre-stored intent recognition model to perform intent recognition on the target text to obtain an intent recognition result. Based on the intent recognition result, the state of the call list is marked as a processed state to provide data point support for subsequent strategy analysis. An index is established for the call list in the processed state to avoid repeated processing.

[0101] Fig.10 Schematic diagram of data processing of the second server provided in the embodiment of the present application. Fig.10 As shown, the second server 40 includes an intelligent outbound call platform, a timing scheduling platform, a database, and a processor. The program code corresponding to the running intention recognition model is pre-stored in the second server 40. During the execution of the slave processor, the intent recognition model is called by loading these program codes. In actual application scenarios, the second server 40 can also deploy other algorithm models, such as emotion recognition models, which are not limited in this application.

[0102] like Fig.10 As shown, the intelligent outbound call platform stores the first text carrying the call identifier in the end sentence compensation list of the database. The timing scheduling platform polls the database based on a preset period to query whether the data stored in the preset time period is the end sentence compensation list. In one example, the preset time period can be in hours. For example, if the current query time is 9 o'clock, the data stored between 8 o'clock and 9 o'clock can be queried.

[0103] In some embodiments of the present application, after the timing scheduling platform obtains data for a preset time period from the database, the data is sent to the processor for analysis. The processor obtains the call identifier in the tail sentence compensation list, and searches the call list in the database based on the call identifier. The subtext of the last round of call speech in the call list is obtained to determine whether the subtext is the voice text corresponding to the target user. If the subtext is the voice text corresponding to the target user, the first text is spliced ​​with the subtext. The spliced ​​text is stored in the first list area where the subtext is located. If the subtext is not the voice text corresponding to the target user, a second list area is created in the call list. The first text is stored in the second list area.

[0104] After the processor determines the storage area of ​​the first text, the intent recognition model is called to perform intent recognition on the target text (including the first text and the second text) to obtain the intent recognition result. Based on the intent recognition result, the state of the call list is marked as processed. The intent recognition result and the call list marked as processed are stored in the database. The database updates the original call list with the call list marked as processed, and stores it in association with the intent recognition result to provide data support for subsequent strategy analysis.

[0105] Fig.11 1 is a schematic diagram of the structure of the speech processing device provided in the embodiment of the present application. The speech processing device 110 may include a plurality of functional modules composed of computer program segments. The computer program of each program segment in the speech processing device 110 may be stored in the memory of the server and executed by at least one processor.

[0106] The receiving module 1101 is used to receive the conversation voice sent by the telephone exchange device.

[0107] The communication link establishing module 1102 is used for, upon receiving a call ending instruction sent by the telephone exchange device, if it is determined that the first text of the call voice has not been sent to the telephone exchange device, sending a connection request to the second server and establishing a communication link between the first server and the second server, wherein the telephone exchange device is used to send the first text to the second server.

[0108] The sending module 1103 is used to send the first text to the second server through the communication link.

[0109] The module referred to in this application refers to a series of computer program segments that can be executed by at least one processor and can complete fixed functions, which are stored in the memory of the server. In this embodiment, the definition of the voice processing device 110 can refer to the definition of the communication compensation method above, and will not be repeated here.

[0110] Fig.12 1 is a schematic diagram of the structure of a speech processing device provided by another embodiment of the present application. The speech processing device 120 may include multiple functional modules composed of computer program segments. The computer program of each program segment in the speech processing device 120 may be stored in a memory of a computer device and executed by at least one processor.

[0111] The first receiving module 1201 is used to establish a communication link between the second server and the first server when receiving a call ending instruction sent by the telephone exchange device and a connection request sent by the first server.

[0112] The second receiving module 1202 is configured to receive a first text sent by the first server through the communication link.

[0113] The module referred to in this application refers to a series of computer program segments that can be executed by at least one processor and can perform fixed functions, which are stored in the memory of the computer device. In this embodiment, the definition of the voice processing device 120 can refer to the definition of the communication compensation method above, and will not be repeated here.

[0114] Fig.13 1 is a schematic diagram of the structure of a computer device provided in an embodiment of the present application. The computer device 130 may be as follows: Figure 1 The second server 40 in the embodiment may also be Figure 1 The computer device 130 includes, but is not limited to, a communication module 1301, a memory 1302, a processor 1303, an input / output (I / O) interface 1304, and a bus 1305. The processor 1303 is coupled to the communication module 1301, the memory 1302, and the I / O interface 1304 through the bus 1305.

[0115] The communication module 1301 may include a wired communication module and / or a wireless communication module. The wired communication module may provide one or more wired communication solutions such as universal serial bus (USB), controller area network bus (CAN, Controller Area Network). The wireless communication module may provide one or more wireless communication solutions such as wireless fidelity (wireless fidelity, Wi-Fi), bluetooth (bluetooth, BT), mobile communication network, frequency modulation (frequency modulation, FM), near field communication technology (near field communication, NFC), infrared technology (infrared, IR), etc.

[0116] The memory 1302 may include one or more random access memories (RAM) and one or more non-volatile memories (NVM). The random access memory may be directly read and written by the processor 1303, and may be used to store executable programs (such as machine instructions) of an operating system or other running programs, and may also be used to store user and application data. The random access memory may include static random-access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), etc.

[0117] The non-volatile memory may also store executable programs and user and application data, etc., and may be loaded into the random access memory in advance for direct reading and writing by the processor 110. The non-volatile memory may include a disk storage device and a flash memory.

[0118] The memory 1302 is used to store one or more computer programs. The one or more computer programs are configured to be executed by the processor 1303. The one or more computer programs include multiple instructions. When the multiple instructions are executed by the processor 1303, the speech processing method executed on the computer device 130 can be implemented.

[0119] In other embodiments, the computer device 130 further includes an external memory interface for connecting to an external memory to expand the storage capacity of the computer device 130 .

[0120] The processor 1303 may include one or more processing units, for example, the processor 1303 may include an application processor (AP), a modem processor, a graphics processor (GPU), an image signal processor (ISP), a controller, a video codec, a digital signal processor (DSP), a baseband processor, and / or a neural-network processing unit (NPU), etc. Different processing units may be independent devices or integrated into one or more processors.

[0121] The processor 1303 provides computing and control capabilities. For example, the processor 1303 is used to execute computer programs stored in the memory 1302 .

[0122] The I / O interface 1304 is used to provide a channel for user input or output. For example, the I / O interface 1304 can be used to connect various input and output devices, such as a mouse, keyboard, touch device, display screen, etc., so that the user can enter information or visualize information.

[0123] The bus 1305 is at least used to provide a channel for mutual communication among the communication module 1301 , the memory 1302 , the processor 1303 , and the I / O interface 1304 in the computer device 130 .

[0124] It is to be understood that the structure illustrated in the embodiment of the present application does not constitute a specific limitation on the computer device 130. In other embodiments of the present application, the computer device 130 may include more or fewer components than shown in the figure, or combine some components, or split some components, or arrange the components differently. The components shown in the figure may be implemented in hardware, software, or a combination of software and hardware.

[0125] The embodiment of the present application also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, each process of the above-mentioned data extraction method embodiment or each process of the above-mentioned data extraction method embodiment is implemented, and the same technical effect can be achieved. To avoid repetition, it is not repeated here. Among them, the computer-readable storage medium is, such as a read-only memory (Read-Only Memory, referred to as ROM), a random access memory (Random Access Memory, referred to as RAM), a disk or an optical disk, etc.

[0126] The embodiment of the present application also provides a computer program product or a computer program, which includes a computer instruction stored in a computer-readable storage medium. The processor of the server reads the computer instruction from the computer-readable storage medium, and the processor executes the computer instruction, so that the server executes the data extraction method described above in the embodiment of the present application.

[0127] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or device including the element.

[0128] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, a magnetic disk, or an optical disk), and includes a number of instructions for a terminal (which can be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to execute the methods described in each embodiment of the present application.

[0129] The embodiments of the present application are described above in conjunction with the accompanying drawings, but the present application is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present application, ordinary technicians in this field can also make many forms without departing from the purpose of the present application and the scope of protection of the claims, all of which are within the protection of the present application.

Claims

1. A voice processing method, applied to a first server, characterized in that: The method comprises: Receiving the conversation voice sent by the telephone exchange device; Upon receiving the call end instruction sent by the telephone exchange device, if it is determined that the first text of the call speech is not sent to the telephone exchange device, sending a connection request to the second server and establishing a communication link between the first server and the second server; The first text is sent to the second server via the communication link.

2. The speech processing method according to claim 1, characterized in that: After receiving the call voice sent by the telephone exchange device, the method further includes: In response to the voice detection instruction sent by the telephone exchange device, the call voice is sent to the audio processing device.

3. The speech processing method according to claim 2, characterized in that: After receiving the call ending instruction sent by the telephone exchange device, the method further includes: If a stop detection instruction sent by the telephone exchange device is not received before the call end instruction is received, it is determined that the first text is not sent to the telephone exchange device; After determining that the first text is not sent to the telephone exchange device, the method further includes: Sending a feedback instruction to the audio processing device; The first text sent by the audio processing device in response to the return instruction is received.

4. The speech processing method according to claim 2, characterized in that: Before receiving the call ending instruction sent by the telephone exchange device, the method further includes: If a stop detection instruction sent by the telephone exchange device is received before the call end instruction is received, and the first text fed back by the audio processing device is not received, it is determined that the first text is not sent to the telephone exchange device.

5. A voice processing method, applied to a second server, characterized in that: include: Upon receiving a call end instruction sent by the telephone exchange device and a connection request sent by the first server, establishing a communication link between the second server and the first server; A first text sent by the first server is received through the communication link.

6. The speech processing method according to claim 5, characterized in that: The first text includes a call identifier, and the method includes: searching a subtext of a second text having the call identifier from the call list, wherein the second text is a text of a call voice received by the telephone exchange device before establishing the communication link, and the subtext is a text corresponding to the last round of call voice in the call list; The subtext of the second text is stored in the call list in correspondence with the first text.

7. The speech processing method according to claim 6, characterized in that: The storing the subtext of the second text and the first text in the call list in correspondence with each other includes: In a case where the subtext is a text of a call speech corresponding to the target user, determining the subtext to be in a first list area in the call list; concatenating the subtext with the first text to obtain a concatenated text; The concatenated text is stored in the first list area.

8. The speech processing method according to claim 6, characterized in that: The storing the subtext of the second text and the first text in the call list in correspondence with each other includes: In the case where the subtext is a text of a call speech corresponding to a non-target user, creating a second list area in the call list; The first text is stored in the second list area.

9. A voice processing device, applied to a first server, characterized in that: include: A receiving module, used for receiving the conversation voice sent by the telephone exchange device; a communication link establishing module, configured to, upon receiving a call ending instruction sent by the telephone exchange device, send a connection request to a second server and establish a communication link between the first server and the second server if it is determined that the first text of the call speech has not been sent to the telephone exchange device; A sending module is used to send the first text to the second server through the communication link.

10. A voice processing device, applied to a second server, characterized in that: include: A first receiving module, configured to establish a communication link between the second server and the first server upon receiving a call ending instruction sent by the telephone exchange device and a connection request sent by the first server; The second receiving module is used to receive the first text sent by the first server through the communication link.

11. A computer device, characterized in that: The computer device includes a processor and a memory, and the processor is used to execute a computer program stored in the memory to implement the speech processing method as described in any one of claims 1 to 4, or to implement the speech processing method as described in any one of claims 5 to 8.

12. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores at least one instruction, and when the at least one instruction is executed by the processor, it implements the speech processing method as described in any one of claims 1 to 4, or implements the speech processing method as described in any one of claims 5 to 8.

13. A computer program product, characterized in that The computer program product stores instructions, which, when executed on a computer, enable the computer to implement the speech processing method as claimed in any one of claims 1 to 4, or the speech processing method as claimed in any one of claims 5 to 8.

Images