Debt collection method and system based on emotion recognition, computer equipment, and storage medium

Abstract

The embodiments of the present application provide a collection method and system based on emotion recognition, a computer device, and a storage medium, which belong to the field of artificial intelligence technology. The method includes: establishing a first call connection between a target collection object and a first service object according to a preset first overdue time interval to obtain first call voice data; obtaining first voice data in the first call voice data; performing risk analysis based on historical collection data, a preset behavior keyword library, and a preset behavior voiceprint feature database, and when the first risk judgment result indicates that the target collection object has no preset behavior, performing emotion recognition on the first voice data; obtaining first voice text data, matching the first voice text data with a collection strategy, and the second service object performs collection processing on the target collection object according to the obtained collection strategy information. The embodiments of the present application can promptly identify and process the user's dissatisfaction, thereby completing the collection task more effectively.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and in particular to a collection method and system based on emotion recognition, a computer device, and a storage medium. Background Art

[0002] At present, in the process of collecting debts from users, AI outbound calling devices or debt collectors are usually used to collect debts from users. However, during the process of debt collection communication with users, debt collectors are easily affected by subjective emotional factors, which can easily lead to words that make users feel dissatisfied. Therefore, when communicating with users on the phone, how to promptly identify users' dissatisfaction and complete the debt collection task more effectively has become a technical problem that needs to be solved urgently. Summary of the invention

[0003] The main purpose of the embodiments of the present application is to propose a collection method and system, computer equipment, and storage medium based on emotion recognition, which can promptly identify and process user dissatisfaction, thereby completing the collection task more effectively.

[0004] To achieve the above-mentioned purpose, a first aspect of an embodiment of the present application proposes a debt collection method based on emotion recognition, the method comprising:

[0005] Establishing a first call connection between the target collection object and the first service object according to a preset first overdue time interval to obtain first call voice data between the target collection object and the first service object;

[0006] Acquire the first voice data of the target debt collection object in the first call voice data;

[0007] Obtaining historical collection data of the target collection object;

[0008] Performing risk analysis on the first voice data according to the historical collection data, the preset behavior keyword library, and the preset behavior voiceprint feature database to obtain a first risk judgment result;

[0009] When the first risk judgment result indicates that the target debt collection object has no preset behavior, emotion recognition is performed on the first voice data according to a preset emotion recognition model to obtain a first emotion recognition result;

[0010] Acquire first voice text data corresponding to the first voice data, and input the first emotion recognition result and the first voice text data into a preset collection strategy model for matching to obtain collection strategy information;

[0011] The collection strategy information is sent to a second service object, and the second service object performs collection processing on the target collection object according to the collection strategy information.

[0012] In some embodiments, after performing risk analysis on the first voice data according to the historical collection data of the target collection object, the preset behavior keyword library, and the preset behavior voiceprint feature database to obtain a first risk judgment result, the method further includes:

[0013] When the first risk judgment result indicates that the target debt collection object has a preset behavior, first voice text data corresponding to the first voice data is obtained, and the first voice text data is input into a preset speech recommendation model for speech recommendation to obtain a target recommended speech;

[0014] The target recommendation words are sent to a third service object.

[0015] In some embodiments, the sending of the collection strategy information to the second service object, and the second service object performing collection processing on the target collection object according to the collection strategy information, includes:

[0016] When the collection completion information of the target collection object is not obtained within the first overdue time interval, sending the collection strategy information to the second service object;

[0017] Establishing a second call connection between the target collection object and the second service object at a preset second overtime time interval according to the collection strategy information to obtain second call voice data between the target collection object and the second service object, wherein the second overtime time interval is used to represent a preset time interval after the end of the first overtime time interval;

[0018] Acquire the second voice data of the target debt collection object in the second call voice data;

[0019] Performing risk analysis on the second voice data according to the historical collection data, the preset behavior keyword library, and the preset behavior voiceprint feature database to obtain a second risk judgment result;

[0020] When the second risk judgment result indicates that the target debt collection object has no preset behavior, emotion recognition is performed on the second voice data according to the emotion recognition model to obtain a second emotion recognition result;

[0021] Acquire second voice text data corresponding to the second voice data, and input the second emotion recognition result and the second voice text data into a preset collection speech model for speech matching to obtain target collection speech;

[0022] Debt collection is performed on the target debt collection object according to the target debt collection script.

[0023] In some embodiments, after performing risk analysis on the second voice data according to the historical collection data, the preset behavior keyword library, and the preset behavior voiceprint feature database to obtain a second risk judgment result, the method further includes:

[0024] Acquire third voice data of the second service object in the second call voice data;

[0025] Performing an emotional state analysis on the third voice data according to a preset emotional state keyword library and an emotional state voiceprint feature database to obtain a collection emotion judgment result;

[0026] Acquire second voice text data corresponding to the second voice data, and input the second voice text data into a preset speech recommendation model for speech recommendation to obtain target recommended speech;

[0027] When the collection emotion judgment result indicates that the second service object is in a preset emotional state, the target recommendation words are sent to the third service object. The third service object collects the target collection object according to the target recommendation words and processes the second service object according to preset management specifications.

[0028] In some embodiments, after performing emotional state analysis on the third voice data according to the preset emotional state keyword library and emotional state voiceprint feature database to obtain the collection emotion judgment result, the method further includes:

[0029] When the collection emotion judgment result indicates that the second service object is not in the preset emotional state, the second service object collects debts from the target collection object according to the target recommended speech.

[0030] In some embodiments, the risk analysis of the first voice data is performed according to the historical collection data, the preset behavior keyword library, and the preset behavior voiceprint feature database to obtain a first risk judgment result, including:

[0031] Acquire first voice text data corresponding to the first voice data;

[0032] Acquire a preset behavior keyword library and a preset behavior voiceprint feature database, wherein the preset behavior keyword library includes a plurality of preset behavior tendency keywords;

[0033] Performing keyword matching processing on the first voice text data and the preset behavior tendency keyword to obtain a preset behavior keyword matching result;

[0034] Obtaining a target voiceprint feature corresponding to the first voice data;

[0035] Performing voiceprint comparison on the target voiceprint feature and the preset behavior voiceprint feature database to obtain a voiceprint comparison result;

[0036] A risk analysis is performed on the first voice data according to the historical collection data, the preset behavior keyword matching result and the voiceprint comparison result to obtain a first risk judgment result.

[0037] In some embodiments, after obtaining first voice text data corresponding to the first voice data, and inputting the first emotion recognition result and the first voice text data into a preset collection strategy model for matching, and obtaining collection strategy information, the method further includes:

[0038] Acquire text features of the first voice text data and voice features of the first voice data;

[0039] Performing intention recognition on the target debt collection object according to the text features and the voice features to obtain an intention recognition result;

[0040] The target collection object is marked according to the intention recognition result to obtain a marking result, and the marking result is used to characterize key intention information of the target collection object in the first voice data.

[0041] To achieve the above-mentioned purpose, a second aspect of an embodiment of the present application proposes a collection system based on emotion recognition, the system comprising:

[0042] A call connection module, used to establish a first call connection between a target collection object and a first service object according to a preset first overdue time interval, so as to obtain first call voice data between the target collection object and the first service object;

[0043] A first voice data acquisition module, used to acquire first voice data of the target collection object in the first call voice data;

[0044] A second voice data acquisition module is used to acquire the historical collection data of the target collection object;

[0045] a risk analysis module, configured to perform risk analysis on the first voice data according to the historical collection data, the preset behavior keyword library, and the preset behavior voiceprint feature database to obtain a first risk judgment result;

[0046] an emotion recognition module, configured to perform emotion recognition on the first voice data according to a preset emotion recognition model to obtain a first emotion recognition result when the first risk judgment result indicates that the target collection object has no preset behavior;

[0047] a matching module, configured to obtain first voice text data corresponding to the first voice data, and input the first emotion recognition result and the first voice text data into a preset collection strategy model for matching to obtain collection strategy information;

[0048] The collection module is used to send the collection strategy information to the second service object, and the second service object performs collection processing on the target collection object according to the collection strategy information.

[0049] To achieve the above objective, a third aspect of an embodiment of the present application provides a computer device, including:

[0050] at least one memory;

[0051] at least one processor;

[0052] at least one computer program;

[0053] The computer program is stored in the memory, and the processor executes at least one computer program to implement the method described in any one of the embodiments of the first aspect.

[0054] To achieve the above-mentioned purpose, the fourth aspect of the embodiment of the present application further proposes a storage medium,

[0055] The storage medium is a computer-readable storage medium, and the computer-readable storage medium stores a computer program, and the computer program is used to enable a computer to execute the method described in any one of the embodiments of the first aspect above.

[0056] The collection method and system based on emotion recognition, computer equipment, and storage medium proposed in the embodiment of the present application, first, establish a first call connection between the target collection object and the first service object according to a preset first overdue time interval to obtain the first call voice data of the target collection object and the first service object. In order to timely and accurately identify the user's dissatisfaction, the first voice data of the target collection object and the historical collection data of the target collection object in the first call voice data are obtained, and then the first voice data is risk analyzed according to the historical collection data, the preset behavior keyword library, and the preset behavior voiceprint feature database to obtain a first risk judgment result. When the first risk judgment result indicates that the target collection object has no preset behavior, the first voice data is emotion recognized according to the preset emotion recognition model to obtain the first emotion recognition result. In order to complete the collection task more efficiently, the first voice text data corresponding to the first voice data is obtained, and the first emotion recognition result and the first voice text data are input into the preset collection strategy model for matching to obtain the collection strategy information. The collection strategy information is sent to the second service object so that the second service object performs collection processing on the target collection object according to the collection strategy information. The embodiments of the present application can promptly identify user dissatisfaction and handle it flexibly, thereby completing the collection task more effectively. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 is a first flow chart of a collection method based on emotion recognition provided in an embodiment of the present application;

[0058] Figure 2 is a second flow chart of the debt collection method based on emotion recognition provided in an embodiment of the present application;

[0059] Figure 3 yes Figure 1 Flow chart of the specific method of step S170;

[0060] Figure 4 is a third flow chart of the debt collection method based on emotion recognition provided in an embodiment of the present application;

[0061] Figure 5 yes Figure 1 Flow chart of the specific method of step S140;

[0062] Figure 6 is a fourth flow chart of the debt collection method based on emotion recognition provided in an embodiment of the present application;

[0063] Figure 7 It is a module block diagram of a collection system based on emotion recognition provided in an embodiment of the present application;

[0064] Figure 8 It is a schematic diagram of the hardware structure of the computer device provided in the embodiment of the present application. DETAILED DESCRIPTION

[0065] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0066] It should be noted that, although the functional modules are divided in the device schematic diagram and the logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than the module division in the device or the order in the flowchart. The terms "first", "second", etc. in the specification, claims and the above drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence.

[0067] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used herein are only for the purpose of describing the embodiments of this application and are not intended to limit this application.

[0068] First, some nouns involved in this application are analyzed:

[0069] Artificial Intelligence (AI): It is a new technical science that studies and develops theories, methods, technologies and application systems for simulating, extending and expanding human intelligence. AI is a branch of computer science. AI attempts to understand the essence of intelligence and produce a new intelligent machine that can respond in a similar way to human intelligence. Research in this field includes robots, language recognition, image recognition, natural language processing and expert systems. AI can simulate the information process of human consciousness and thinking. AI is also a theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

[0070] Deep Neural Networks-Hidden Markov Model (DNN-HMM): refers to using DNN to replace the Gaussian Mixture Model (GMM) to model the observation probability of the input speech signal. GMM is a generative model that uses unsupervised learning, while DNN is a discriminative model that uses supervised learning.

[0071] At present, in the process of collecting debts from users, AI outbound calling devices or debt collectors are usually used to collect debts from users. However, since AI outbound calling devices usually use a unified speech template to reply, it is easy for users to feel dissatisfied because they cannot get the desired reply, and debt collectors are easily affected by subjective emotional factors during the debt collection communication with users, which can easily lead to the words they say causing dissatisfaction among users. Therefore, when communicating with users on the phone, how to promptly identify users' dissatisfaction and accurately navigate users to the corresponding customer service personnel or self-service processes, so as to more effectively complete the debt collection task, has become a technical problem that needs to be solved urgently.

[0072] Based on this, the embodiments of the present application provide a collection method and system based on emotion recognition, a computer device, and a storage medium, which can promptly identify and process the user's dissatisfaction, thereby completing the collection task more effectively.

[0073] The collection method and system based on emotion recognition, computer equipment, and storage medium provided in the embodiments of the present application are specifically illustrated through the following embodiments. First, the collection method based on emotion recognition in the embodiments of the present application is described.

[0074] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Among them, artificial intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

[0075] AI basic technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing technology, operation / interaction systems, mechatronics, etc. AI software technologies mainly include computer vision technology, robotics technology, biometrics technology, speech processing technology, natural language processing technology, and machine learning / deep learning.

[0076] The collection method based on emotion recognition provided in the embodiment of the present application relates to the field of artificial intelligence technology, and in particular to the field of data mining technology. The collection method based on emotion recognition provided in the embodiment of the present application can be applied to a terminal, or to a server, or can be software running in a terminal or a server. In some embodiments, the terminal can be a smart phone, a tablet computer, a laptop computer, a desktop computer, or a smart watch, etc.; the server can be an independent server, or it can be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (Content Delivery Network, CDN), and big data and artificial intelligence platforms; the software can be an application that implements a collection method based on emotion recognition, etc., but is not limited to the above forms.

[0077] The present application can be used in many general or special computer system environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, etc. The present application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. The present application can also be practiced in distributed computing environments, in which tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules can be located in local and remote computer storage media including storage devices.

[0078] Please refer to Figure 1 , Figure 1 is an optional flow chart of the debt collection method based on emotion recognition provided in the embodiment of the present application. In some embodiments of the present application, the debt collection method based on emotion recognition specifically includes but is not limited to steps S110 to S170. Figure 1 These seven steps are introduced in detail.

[0079] Step S110, establishing a first call connection between the target collection object and the first service object according to a preset first overdue time interval to obtain first call voice data between the target collection object and the first service object;

[0080] Step S120, obtaining first voice data of the target collection object in the first call voice data;

[0081] Step S130, obtaining historical collection data of the target collection object;

[0082] Step S140, performing risk analysis on the first voice data according to the historical collection data, the preset behavior keyword library, and the preset behavior voiceprint feature database to obtain a first risk judgment result;

[0083] Step S150, when the first risk judgment result indicates that the target collection object has no preset behavior, emotion recognition is performed on the first voice data according to a preset emotion recognition model to obtain a first emotion recognition result;

[0084] Step S160, obtaining first voice text data corresponding to the first voice data, and inputting the first emotion recognition result and the first voice text data into a preset collection strategy model for matching to obtain collection strategy information;

[0085] Step S170: Send the collection strategy information to the second service object, and the second service object performs collection processing on the target collection object according to the collection strategy information.

[0086] In step S110 of some embodiments, since the user has not repaid the loan after the due date T, it is necessary to collect the user, and a first call connection is established between the target collection object and the first service object according to the preset first overdue time interval to obtain the first call voice data of the target collection object and the first service object. Among them, n1 can represent the first overdue time interval, and the time interval of the first overdue time interval can be set to (T, T+n1). Specifically, within the time interval of the first overdue time interval, a communication connection is established between the first service object and the target collection object to obtain the first call voice data of the target collection object and the first service object, and the first call voice data is used to represent the interactive audio stream data between the target collection object and the first service object.

[0087] It should be noted that in order to reduce manpower consumption and improve collection efficiency, the first service object can use AI outbound calls for collection notifications.

[0088] In step S120 of some embodiments, in order to timely judge the user's emotions, first voice data of the target collection object in the first call voice data is obtained, and the first voice data is used to represent the voice data input by the target collection object in the first call voice data.

[0089] In step S130 of some embodiments, in order to accurately identify whether the target collection object has a preset behavior, the historical collection data of the target collection object is obtained. The historical collection data includes historical call voice data, historical preset behavior keyword statistics, historical preset behavior occurrence statistics, collection record data, etc. The historical collection process may include at least one collection process. The historical preset behavior keyword statistics are used to characterize the statistical results of the keywords of the preset behavior of the target collection object in the historical collection process; the historical call voice data are used to characterize the interactive audio stream data of the historical communication between the target collection object and the first service object; the historical preset behavior occurrence statistics are used to characterize the number of times the target collection object actually has the preset behavior in the historical collection process; the collection record data are used to characterize the time, frequency, and other records of the target collection object being collected. It should be noted that, for example, in the credit collection process of a bank, the historical collection data can be obtained from the bank's cloud server.

[0090] It should be noted that the preset behavior includes at least one behavioral intention in the target collection object's voice, such as clear expression of dissatisfaction, unhealthy words, intense emotions, etc. Data statistics are collected based on the preset behavior keywords corresponding to different behavioral intentions to form historical collection data.

[0091] In step S140 of some embodiments, in order to accurately obtain a risk judgment on the target collection object, the first voice data is subjected to a risk analysis based on the historical collection data, the preset behavior keyword library, and the preset behavior voiceprint feature database to obtain a first risk judgment result. The preset behavior keyword library is used to store multiple preset behavior tendency keywords under different business categories, and the preset behavior keyword library is combined to determine whether the text data corresponding to the first voice data contains keywords expressing the preset behavior intention, so as to obtain a keyword judgment result. The preset behavior voiceprint feature database is used to store voiceprint features that can express preset behaviors, and the preset behavior voiceprint feature database is combined to determine whether the voiceprint features in the first voice data meet the voiceprint threshold of the preset behavior, so as to obtain a voiceprint judgment result. Specifically, the state of the current target collection object is subjected to a risk analysis by combining the keyword judgment result, the voiceprint judgment result, and the historical collection data, and its corresponding emotion level is determined, which is the first risk judgment result.

[0092] It should be noted that the emotion levels corresponding to the risk judgment results can be divided into level one, level two and level three. When the emotion level of the target collection object is at level two, it indicates that the target collection object may currently be in a relatively stable and normal emotion, that is, there may be no risk of preset behavior; when the emotion level of the target collection object is at level one or level three, it indicates that the target collection object may currently be in a low or intense emotion, that is, there may be a risk of preset behavior.

[0093] In step S150 of some embodiments, when the first risk judgment result indicates that the target collection object has no preset behavior, in order to be able to timely and accurately judge the user's emotions and process them, emotion recognition is performed on the first voice data according to a preset emotion recognition model to obtain a first emotion recognition result. Specifically, when the first risk judgment result indicates that the target collection object has no preset behavior, audio features are extracted from the first voice data to obtain a first audio feature vector. Then, the first audio feature vector is input into a preset emotion recognition model and emotion recognition is performed to obtain the current emotion category of the target collection object, that is, the first emotion recognition result. Among them, emotion categories include: calm, intense, frustrated, etc., and intense emotions include anger, boredom, etc.

[0094] It should be noted that the preset emotion recognition model can adopt any one of the DNN-HMM acoustic model, N-Gram language model or emotion analysis model based on finite weighted state transition machine, and the present application does not limit the emotion recognition model.

[0095] In step S160 of some embodiments, in order to effectively improve the completion of the collection task, ensure the smooth communication between the target collection object and the service object, and avoid the dissatisfaction of the target object, the first voice text data corresponding to the first voice data is obtained, and the first emotion recognition result and the first voice text data are input into the preset collection strategy model for matching, so as to obtain the collection strategy information through the collection strategy model. The collection strategy information includes collection frequency recommendation information, collection recommended speech information, etc.

[0096] In step S170 of some embodiments, in order to improve the completion of the collection task, the collection strategy information is sent to the second service object, and the second service object performs collection processing on the target collection object according to the collection strategy information. Specifically, the second service object can collect the target collection object according to the collection recommendation speech information and the collection frequency suggestion information, thereby improving the success rate of the collection task. It should be noted that the second service object can be an AI outbound call or a collection person.

[0097] It should be noted that in each specific implementation of the present application, when it comes to the need to perform relevant processing based on data related to user identity or characteristics such as user information, user behavior data, user historical data, and user location information, the user's permission or consent will be obtained first, and the collection, use, and processing of these data will comply with the relevant laws, regulations, and standards of the relevant countries and regions. In addition, when the embodiment of the present application needs to obtain the user's sensitive personal information, the user's separate permission or consent will be obtained through a pop-up window or by jumping to a confirmation page. After clearly obtaining the user's separate permission or consent, the necessary user-related data for the normal operation of the embodiment of the present application will be obtained.

[0098] Please refer to Figure 2 , Figure 2 is another optional flow chart of the debt collection method based on emotion recognition provided by the embodiment of the present application. In some embodiments of the present application, after step S140, the debt collection method based on emotion recognition specifically further includes but is not limited to step S210 and step S220. Figure 2 These two steps are introduced in detail.

[0099] Step S210: When the first risk judgment result indicates that the target collection object has a preset behavior, first voice text data corresponding to the first voice data is obtained, and the first voice text data is input into a preset speech recommendation model for speech recommendation to obtain a target recommended speech;

[0100] Step S220: sending the target recommendation words to the third service object.

[0101] In step S210 and step S220 of some embodiments, when the first risk judgment result indicates that the target collection object has a preset behavior, in order to effectively ensure the completion of the collection task, the first voice text data corresponding to the first voice data is obtained, and the first voice text data is input into the preset speech recommendation model for speech recommendation to obtain the target recommended speech, and the target recommended speech is sent to the third service object. Specifically, when the first risk judgment result indicates that the target collection object has a preset behavior, that is, there is a clear expression of dissatisfaction, unhealthy words, intense emotional expression and at least one behavioral intention in the voice of the target collection object, the first voice text data is input into the preset speech recommendation model for speech recommendation to obtain the target recommended speech. For example, when the first risk judgment result indicates that the target collection object is currently in an intense emotional state, the third service object can appease and express the target collection object according to the target recommended speech, thereby alleviating the emotional state of the target collection object for subsequent calls. The embodiment of the present application can timely identify the user's dissatisfaction and flexibly handle it by judging in real time whether the target collection object has the risk of preset behavior, thereby more effectively ensuring the completion of the collection task.

[0102] Please refer to Figure 3 , Figure 3 is a flowchart of a specific method of step S170 provided in an embodiment of the present application. In some embodiments of the present application, step S170 specifically includes but is not limited to steps S310 to S370. Figure 3 These seven are introduced in detail.

[0103] Step S310, when the collection completion information of the target collection object is not obtained within the first overtime time interval, the collection strategy information is sent to the second service object;

[0104] Step S320, establishing a second call connection between the target collection object and the second service object at a preset second overtime time interval according to the collection strategy information to obtain second call voice data between the target collection object and the second service object, wherein the second overtime time interval is used to represent a preset time interval after the end of the first overtime time interval;

[0105] Step S330, obtaining second voice data of the target collection object in the second call voice data;

[0106] Step S340, performing risk analysis on the second voice data according to the historical collection data, the preset behavior keyword library, and the preset behavior voiceprint feature database to obtain a second risk judgment result;

[0107] Step S350: when the second risk judgment result indicates that the target debt collection object has no preset behavior, emotion recognition is performed on the second voice data according to the emotion recognition model to obtain a second emotion recognition result;

[0108] Step S360, obtaining second voice text data corresponding to the second voice data, and inputting the second emotion recognition result and the second voice text data into a preset collection speech model for speech matching to obtain a target collection speech;

[0109] Step S370, collecting debts from the target collection object according to the target collection words.

[0110] In step S310 and step S320 of some embodiments, when the collection completion information of the target collection object is not obtained in the first overtime interval, that is, the collection task for the target collection object is not completed within the first overtime interval, the collection strategy information is sent to the second service object, so that the second service object performs collection processing on the target collection object according to the collection strategy information. Then, the second service object establishes a second call connection between the target collection object and the second service object in the preset second overtime interval according to the collection strategy information to obtain the second call voice data of the target collection object and the second service object. Among them, the second overtime interval is used to represent the preset time interval after the end of the first overtime interval, n2 can represent the second overtime interval, and the time interval of the second overtime interval can be set to (T+n1, T+n2). Specifically, within the time interval of the second overtime interval, a communication connection is established between the second service object and the target collection object to obtain the second call voice data of the target collection object and the second service object, and the second call voice data is used to represent the interactive audio stream data between the target collection object and the second service object.

[0111] In step S330 and step S340 of some embodiments, in order to timely judge the user's emotions, the second voice data of the target collection object in the second call voice data is obtained, and the second voice data is used to represent the voice data input by the target collection object in the second call voice data. And the second voice data is subjected to risk analysis based on the historical collection data, the preset behavior keyword library and the preset behavior voiceprint feature database to obtain a second risk judgment result. Among them, the specific introduction of the historical collection data, the preset behavior keyword library and the preset behavior voiceprint feature database is the same as that in the above embodiment, and will not be repeated here.

[0112] In steps S350 to S370 of some embodiments, when the second risk judgment result indicates that the target collection object has no preset behavior, in order to be able to timely and accurately judge the user's emotions and process them, the second voice data is subjected to emotion recognition according to the emotion recognition model to obtain a second emotion recognition result. Specifically, when the second risk judgment result indicates that the target collection object has no preset behavior, the second voice data is subjected to audio feature extraction to obtain a second audio feature vector. Then, the second audio feature vector is input into a preset emotion recognition model and emotion recognition is performed to obtain the current emotion category of the target collection object, i.e., the second emotion recognition result. In order to effectively improve the completion of the collection task, ensure the smooth progress of the call between the target collection object and the service object, and avoid the occurrence of dissatisfaction of the target object, the second voice text data corresponding to the second voice data is obtained, and the second emotion recognition result and the second voice text data are input into a preset collection speech model for speech matching to obtain the target collection speech, and the target collection object is collected according to the target collection speech to improve the success rate of the collection task.

[0113] Please refer to Figure 4 , Figure 4 is another optional flow chart of the debt collection method based on emotion recognition provided by the embodiment of the present application. In some embodiments, after step S340, the debt collection method based on emotion recognition specifically includes but is not limited to steps S410 to S440. Figure 4 These four steps are introduced in detail.

[0114] Step S410, obtaining third voice data of the second service object in the second call voice data;

[0115] Step S420, performing emotional state analysis on the third voice data according to a preset emotional state keyword library and an emotional state voiceprint feature database to obtain a collection emotion judgment result;

[0116] Step S430, obtaining second voice text data corresponding to the second voice data, and inputting the second voice text data into a preset speech recommendation model for speech recommendation to obtain target recommended speech;

[0117] Step S440, when the collection emotion judgment result indicates that the second service object is in a preset emotional state, the target recommendation words are sent to the third service object, the third service object collects the target collection object according to the target recommendation words, and processes the second service object according to the preset management specifications.

[0118] In step S410 and step S420 of some embodiments, in order to ensure the smooth conduct of the call between the target collection object and the service object, the third voice data of the second service object in the second call voice data is obtained in real time, and the third voice data is analyzed for the emotional state according to the preset emotional state keyword library and emotional state voiceprint feature database to obtain the collection emotion judgment result. Among them, the emotional state keyword library is used to store keywords under different emotional states, and the emotional state includes calm, intense, depressed, etc., and the intense emotion includes anger, boredom, etc. By combining the emotional state keyword library, it is judged whether the text data corresponding to the third voice data contains the keywords under the emotional state, so as to obtain the keyword emotion judgment result. The emotional state voiceprint feature database is used to store voiceprint features that can express the emotional state, and the emotional state satisfied by the voiceprint features in the third voice data is judged by combining the emotional state voiceprint feature database to obtain the voiceprint emotion judgment result. Afterwards, the collection emotion judgment result is obtained according to the keyword emotion judgment result and the voiceprint emotion judgment result.

[0119] In step S430 and step S440 of some embodiments, the second voice text data corresponding to the second voice data is obtained, and the second voice text data is input into the preset speech recommendation model for speech recommendation to obtain the target recommended speech. When the collection emotion judgment result indicates that the second service object is in a preset emotional state, the target recommended speech is sent to the third service object, wherein the preset emotional state can be set to an intense emotional state, that is, the current intense emotional state of the second service object may cause dissatisfaction in the user, and therefore the target recommended speech is sent to the third service object. The third service object collects the target collection object according to the target recommended speech. In order to ensure that the second service object can avoid intense emotional states in the subsequent collection process, the second service object can be processed according to the preset management specifications, wherein the preset management specifications can be adjusted according to actual regulations and requirements, for example, deducting points from the work attitude of the second service object and teaching it.

[0120] It should be noted that in order to ensure the smooth conduct of the conversation between the target collection object and the service object, after the first call connection is established, the fourth voice data of the first service object in the first call voice data can also be obtained in real time, and the emotional state of the fourth voice data is analyzed according to the preset emotional state keyword library and emotional state voiceprint feature database to obtain the collection emotion judgment result. When the collection emotion judgment result indicates that the first service object is in a preset emotional state, the obtained target recommended words are sent to the third service object. The third service object collects the target collection object according to the target recommended words, and processes the second service object according to the preset management specifications.

[0121] In some embodiments, after step S430, the debt collection method based on emotion recognition further includes:

[0122] When the collection emotion judgment result indicates that the second service object is not in the preset emotional state, the second service object collects debts from the target collection object according to the target recommended speech.

[0123] Specifically, when the collection emotion judgment result indicates that the second service object is not in the preset emotional state, that is, it indicates that the current second service object can still serve the target collection object well, the second service object collects the target collection object according to the target recommendation speech. The embodiment of the present application can timely adjust the third service object to collect the target collection object when it is judged that the second service object is in the preset emotional state, thereby avoiding the user's dissatisfaction and more effectively ensuring the completion of the collection task.

[0124] Please refer to Figure 5 , Figure 5 is a flowchart of step S140 provided in an embodiment of the present application. In some embodiments, step S140 specifically includes but is not limited to steps S510 to S560. Figure 5 These six steps are introduced in detail.

[0125] Step S510, obtaining first voice text data corresponding to the first voice data;

[0126] Step S520, obtaining a preset behavior keyword library and a preset behavior voiceprint feature database, wherein the preset behavior keyword library includes a plurality of preset behavior tendency keywords;

[0127] Step S530, performing keyword matching processing on the first voice text data and the preset behavior tendency keyword to obtain a preset behavior keyword matching result;

[0128] Step S540, obtaining a target voiceprint feature corresponding to the first voice data;

[0129] Step S550, performing voiceprint comparison between the target voiceprint feature and the preset behavior voiceprint feature database to obtain a voiceprint comparison result;

[0130] Step S560: Perform risk analysis on the first voice data according to the historical collection data, the preset behavior keyword matching result and the voiceprint comparison result to obtain a first risk judgment result.

[0131] In steps S510 to S560 of some embodiments, in order to accurately obtain a risk judgment on the target collection object, the first voice text data corresponding to the first voice data is obtained, and a preset behavior keyword library and a preset behavior voiceprint feature database are obtained, and the preset behavior keyword library includes multiple preset behavior tendency keywords. By performing keyword matching processing on the first voice text data and the preset behavior tendency keywords, a preset behavior keyword matching result is obtained. After that, the target voiceprint feature corresponding to the first voice data is obtained, and a voiceprint comparison is performed on the target voiceprint feature and the preset behavior voiceprint feature database to obtain a voiceprint comparison result. Finally, a risk analysis is performed on the first voice data based on the historical collection data, the preset behavior keyword matching result, and the voiceprint comparison result, and the corresponding emotion level is determined, which is the first risk judgment result.

[0132] Please refer to Figure 6 , Figure 6 is another optional flow chart of the debt collection method based on emotion recognition provided by the embodiment of the present application. In some embodiments, after step S160, the debt collection method based on emotion recognition specifically includes but is not limited to steps S610 to S630. Figure 6 These three steps are introduced in detail.

[0133] Step S610, obtaining text features of first voice text data and voice features of first voice data;

[0134] Step S620, performing intention recognition on the target collection object according to the text features and the voice features to obtain an intention recognition result;

[0135] Step S630, marking the target collection object according to the intention recognition result to obtain a marking result, and the marking result is used to characterize the key intention information of the target collection object in the first voice data.

[0136] In some embodiments, in steps S610 to S630, in order to accurately record the intention information of the target collection object in the collection process, the text features of the first voice text data and the voice features of the first voice data are obtained, and the intention of the target collection object is recognized based on the text features and the voice features to obtain the intention recognition result. Finally, the target collection object is marked according to the intention recognition result to obtain the marking result, and the marking result is used to characterize the key intention information of the target collection object in the first voice data.

[0137] It should be noted that during the second call connection process, it is also necessary to obtain the text features corresponding to the second voice text data and the voice features of the first voice data, and perform intent recognition on the target collection object based on the text features and voice features to obtain intent recognition results. Finally, the target collection object is labeled according to the intent recognition result to obtain a labeling result, which is used to characterize the key intent information of the target collection object in the second voice data. The collection method based on emotion recognition in the embodiment of the present application can be applied to the emotional warning process for the collection link, and can promptly identify the user's dissatisfaction and handle it flexibly, thereby completing the collection task more effectively.

[0138] Please refer to Figure 7 , Figure 7 7 is a block diagram of a collection system based on emotion recognition provided in some embodiments of the present application. In some embodiments, the collection system based on emotion recognition includes a call connection module 710, a first voice data acquisition module 720, a second voice data acquisition module 730, a risk analysis module 740, an emotion recognition module 750, a matching module 760 and a collection module 770.

[0139] A call connection module 710, configured to establish a first call connection between the target collection object and the first service object according to a preset first overdue time interval, so as to obtain first call voice data between the target collection object and the first service object;

[0140] The first voice data acquisition module 720 is used to acquire the first voice data of the target collection object in the first call voice data;

[0141] The second voice data acquisition module 730 is used to acquire the historical collection data of the target collection object;

[0142] The risk analysis module 740 is used to perform risk analysis on the first voice data according to the historical collection data, the preset behavior keyword library and the preset behavior voiceprint feature database to obtain a first risk judgment result;

[0143] The emotion recognition module 750 is used to perform emotion recognition on the first voice data according to a preset emotion recognition model to obtain a first emotion recognition result when the first risk judgment result indicates that the target collection object has no preset behavior;

[0144] The matching module 760 is used to obtain the first voice text data corresponding to the first voice data, and input the first emotion recognition result and the first voice text data into a preset collection strategy model for matching to obtain collection strategy information;

[0145] The collection module 770 is used to send the collection strategy information to the second service object, and the second service object performs collection processing on the target collection object according to the collection strategy information.

[0146] It should be noted that the collection system based on emotion recognition in the embodiment of the present application is used to implement the above-mentioned collection method based on emotion recognition. The collection system based on emotion recognition in the embodiment of the present application corresponds to the above-mentioned collection method based on emotion recognition. For the specific processing process, please refer to the above-mentioned collection method based on emotion recognition, which will not be elaborated here one by one.

[0147] The present application also provides a computer device, including:

[0148] at least one memory;

[0149] at least one processor;

[0150] at least one computer program;

[0151] The program is stored in the memory, and the processor executes at least one computer program to implement the collection method based on emotion recognition in the above-mentioned embodiment of the present application. The computer device can be any intelligent terminal including a mobile phone, a tablet computer, a personal digital assistant (PDA), a car computer, etc.

[0152] Combine the following Figure 8 The computer device of the embodiment of the present application is introduced in detail.

[0153] like Figure 8 , Figure 8 The hardware structure of a computer device according to another embodiment is illustrated. The computer device includes:

[0154] The processor 810 may be implemented by a general-purpose central processing unit (CPU), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of the present application;

[0155] The memory 820 can be implemented in the form of a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 820 can store an operating system and other application programs. When the technical solution provided in the embodiment of this specification is implemented by software or firmware, the relevant program code is stored in the memory 820, and the processor 810 calls and executes the collection method based on emotion recognition in the embodiment of this application;

[0156] Input / output interface 830, used to implement information input and output;

[0157] The communication interface 840 is used to realize the communication interaction between the device and other devices. The communication can be realized through a wired manner (such as USB, network cable, etc.) or a wireless manner (such as mobile network, WIFI, Bluetooth, etc.);

[0158] bus 850 , which transmits information between the various components of the device (e.g., processor 810 , memory 820 , input / output interface 830 , and communication interface 840 );

[0159] The processor 810 , the memory 820 , the input / output interface 830 and the communication interface 840 are connected to each other in communication within the device via a bus 850 .

[0160] An embodiment of the present application also provides a storage medium, which is a computer-readable storage medium. The computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are used to enable a computer to execute the collection method based on emotion recognition of the above embodiment.

[0161] The memory, as a non-transient computer-readable storage medium, can be used to store non-transient software programs and non-transient computer executable programs. In addition, the memory may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some embodiments, the memory may optionally include a memory remotely disposed relative to the processor, and these remote memories may be connected to the processor via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0162] The embodiments described in the embodiments of the present application are intended to more clearly illustrate the technical solutions of the embodiments of the present application and do not constitute a limitation on the technical solutions provided in the embodiments of the present application. Those skilled in the art will appreciate that with the evolution of technology and the emergence of new application scenarios, the technical solutions provided in the embodiments of the present application are also applicable to similar technical problems.

[0163] Those skilled in the art will appreciate that the technical solutions shown in the figures do not constitute a limitation on the embodiments of the present application, and may include more or fewer steps than shown in the figures, or a combination of certain steps, or different steps.

[0164] The device embodiments described above are merely illustrative, and the units described as separate components may or may not be physically separated, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0165] Those skilled in the art will appreciate that all or some of the steps in the methods disclosed above, and the functional modules / units in the systems and devices may be implemented as software, firmware, hardware, or a suitable combination thereof.

[0166] The terms "first", "second", "third", "fourth", etc. (if any) in the specification of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0167] It should be understood that in the present application, "at least one (item)" means one or more, and "plurality" means two or more. "And / or" is used to describe the association relationship of associated objects, indicating that three relationships may exist. For example, "A and / or B" can mean: only A exists, only B exists, and A and B exist at the same time, where A and B can be singular or plural. The character " / " generally indicates that the objects associated before and after are in an "or" relationship. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, at least one of a, b or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, c can be single or multiple.

[0168] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0169] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0170] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0171] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including multiple instructions to enable an electronic device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of each embodiment method of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), disk or optical disk and other media that can store programs.

[0172] The preferred embodiments of the present invention are described above with reference to the accompanying drawings, but the scope of the rights of the present invention is not limited thereto. Any modification, equivalent substitution and improvement made by a person skilled in the art without departing from the scope and essence of the present invention should be within the scope of the rights of the present invention.

Claims

1. A debt collection method based on emotion recognition, characterized in that: The method comprises: Establishing a first call connection between the target collection object and the first service object according to a preset first overdue time interval to obtain first call voice data between the target collection object and the first service object; Acquire the first voice data of the target debt collection object in the first call voice data; Obtaining historical collection data of the target collection object; Performing risk analysis on the first voice data according to the historical collection data, the preset behavior keyword library and the preset behavior voiceprint feature database to obtain a first risk judgment result; performing risk analysis on the first voice data according to the historical collection data, the preset behavior keyword library and the preset behavior voiceprint feature database to obtain a first risk judgment result, including: obtaining first voice text data corresponding to the first voice data; obtaining a preset behavior keyword library and a preset behavior voiceprint feature database, the preset behavior keyword library including a plurality of preset behavior tendency keywords; performing keyword matching processing on the first voice text data and the preset behavior tendency keywords to obtain a preset behavior keyword matching result; obtaining a target voiceprint feature corresponding to the first voice data; performing voiceprint comparison on the target voiceprint feature and the preset behavior voiceprint feature database to obtain a voiceprint comparison result; performing risk analysis on the first voice data according to the historical collection data, the preset behavior keyword matching result and the voiceprint comparison result to obtain a first risk judgment result; When the first risk judgment result indicates that the target debt collection object has no preset behavior, emotion recognition is performed on the first voice data according to a preset emotion recognition model to obtain a first emotion recognition result; Acquire first voice text data corresponding to the first voice data, and input the first emotion recognition result and the first voice text data into a preset collection strategy model for matching to obtain collection strategy information; The collection strategy information is sent to a second service object, and the second service object performs collection processing on the target collection object according to the collection strategy information; the collection strategy information is sent to a second service object, and the second service object performs collection processing on the target collection object according to the collection strategy information, including: when the collection completion information of the target collection object is not obtained within the first overtime interval, the collection strategy information is sent to the second service object; according to the collection strategy information, a second call connection is established between the target collection object and the second service object at a preset second overtime interval to obtain the second call voice data between the target collection object and the second service object, wherein the second overtime interval is used to represent the first overtime period. a preset time interval after the end of the interval; obtaining the second voice data of the target collection object in the second call voice data; performing risk analysis on the second voice data according to the historical collection data, the preset behavior keyword library and the preset behavior voiceprint feature database to obtain a second risk judgment result; when the second risk judgment result indicates that the target collection object has no preset behavior, performing emotion recognition on the second voice data according to the emotion recognition model to obtain a second emotion recognition result; obtaining the second voice text data corresponding to the second voice data, and inputting the second emotion recognition result and the second voice text data into the preset collection speech model for speech matching to obtain the target collection speech; collecting from the target collection object according to the target collection speech.

2. The method according to claim 1, characterized in that After performing risk analysis on the first voice data according to the historical collection data of the target collection object, the preset behavior keyword library, and the preset behavior voiceprint feature database to obtain a first risk judgment result, the method further includes: When the first risk judgment result indicates that the target debt collection object has a preset behavior, first voice text data corresponding to the first voice data is obtained, and the first voice text data is input into a preset speech recommendation model for speech recommendation to obtain a target recommended speech; The target recommendation words are sent to a third service object.

3. The method according to claim 1, characterized in that After performing risk analysis on the second voice data according to the historical collection data, the preset behavior keyword library, and the preset behavior voiceprint feature database to obtain a second risk judgment result, the method further includes: Acquire third voice data of the second service object in the second call voice data; Performing an emotional state analysis on the third voice data according to a preset emotional state keyword library and an emotional state voiceprint feature database to obtain a collection emotion judgment result; Acquire second voice text data corresponding to the second voice data, and input the second voice text data into a preset speech recommendation model for speech recommendation to obtain target recommended speech; When the collection emotion judgment result indicates that the second service object is in a preset emotional state, the target recommendation words are sent to the third service object. The third service object collects the target collection object according to the target recommendation words and processes the second service object according to preset management specifications.

4. The method according to claim 3, characterized in that After obtaining the second voice text data corresponding to the second voice data, and inputting the second voice text data into a preset speech recommendation model for speech recommendation, and obtaining a target recommended speech, the method further includes: When the collection emotion judgment result indicates that the second service object is not in the preset emotional state, the second service object collects debts from the target collection object according to the target recommended speech.

5. The method according to any one of claims 1 to 4, characterized in that: After obtaining first voice text data corresponding to the first voice data, and inputting the first emotion recognition result and the first voice text data into a preset collection strategy model for matching, and obtaining collection strategy information, the method further includes: Acquire text features of the first voice text data and voice features of the first voice data; Performing intention recognition on the target debt collection object according to the text features and the voice features to obtain an intention recognition result; The target collection object is marked according to the intention recognition result to obtain a marking result, and the marking result is used to characterize key intention information of the target collection object in the first voice data.

6. A debt collection system based on emotion recognition, characterized in that: The system comprises: A call connection module, used to establish a first call connection between a target collection object and a first service object according to a preset first overdue time interval, so as to obtain first call voice data between the target collection object and the first service object; A first voice data acquisition module, used to acquire first voice data of the target collection object in the first call voice data; A second voice data acquisition module is used to acquire the historical collection data of the target collection object; a risk analysis module, configured to perform risk analysis on the first voice data according to the historical collection data, the preset behavior keyword library, and the preset behavior voiceprint feature database to obtain a first risk judgment result; the risk analysis on the first voice data according to the historical collection data, the preset behavior keyword library, and the preset behavior voiceprint feature database to obtain a first risk judgment result, including: obtaining first voice text data corresponding to the first voice data; obtaining a preset behavior keyword library and a preset behavior voiceprint feature database, wherein the preset behavior keyword library includes a plurality of preset behavior tendency keywords; performing keyword matching processing on the first voice text data and the preset behavior tendency keywords to obtain a preset behavior keyword matching result; obtaining a target voiceprint feature corresponding to the first voice data; performing voiceprint comparison between the target voiceprint feature and the preset behavior voiceprint feature database to obtain a voiceprint comparison result; performing risk analysis on the first voice data according to the historical collection data, the preset behavior keyword matching result, and the voiceprint comparison result to obtain a first risk judgment result; an emotion recognition module, configured to perform emotion recognition on the first voice data according to a preset emotion recognition model to obtain a first emotion recognition result when the first risk judgment result indicates that the target collection object has no preset behavior; a matching module, configured to obtain first voice text data corresponding to the first voice data, and input the first emotion recognition result and the first voice text data into a preset collection strategy model for matching to obtain collection strategy information; A collection module, used for sending the collection strategy information to a second service object, and the second service object performs collection processing on the target collection object according to the collection strategy information; sending the collection strategy information to the second service object, and the second service object performs collection processing on the target collection object according to the collection strategy information, including: when the collection completion information of the target collection object is not obtained within the first overtime time interval, sending the collection strategy information to the second service object; establishing a second call connection between the target collection object and the second service object at a preset second overtime time interval according to the collection strategy information, so as to obtain the second call voice data between the target collection object and the second service object, wherein the second overtime time interval is used to represent the collection completion information of the target collection object within the first a preset time interval after the end of the overdue time interval; obtaining the second voice data of the target collection object in the second call voice data; performing risk analysis on the second voice data according to the historical collection data, the preset behavior keyword library and the preset behavior voiceprint feature database to obtain a second risk judgment result; when the second risk judgment result indicates that the target collection object has no preset behavior, performing emotion recognition on the second voice data according to the emotion recognition model to obtain a second emotion recognition result; obtaining the second voice text data corresponding to the second voice data, and inputting the second emotion recognition result and the second voice text data into the preset collection speech model for speech matching to obtain the target collection speech; collecting the target collection object according to the target collection speech.

7. A computer device, characterized in that: include: at least one memory; at least one processor; at least one computer program; The computer program is stored in the memory, and the processor executes the at least one computer program to implement: The method according to any one of claims 1 to 5.

8. A storage medium, the storage medium being a computer-readable storage medium, characterized in that: The computer readable storage medium stores a computer program, and the computer program is used to cause a computer to execute: The method according to any one of claims 1 to 5.

Images