Service abnormity early warning method and device based on voiceprint technology
By applying voiceprint technology in the insurance industry, collecting and comparing voiceprint features, identifying sensitive words and customer emotions, the problems of customer identity identification and malicious complaint handling in the insurance industry are solved, the identification and processing efficiency is improved, and more refined services are provided.
Patent Information
- Application Number
- CN202510318313.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-03-18
- Publication Date
- 2025-06-13
AI Technical Summary
In the insurance industry, in the complaint application, insurance claims and insurance refund, it is impossible to accurately identify the identity of the customer, and the recognition and handling efficiency of malicious complaints are inefficient.
The business abnormal warning method based on voiceprint technology is adopted. By collecting voiceprint features and customer information, we can determine whether there are sensitive words in the voiceprint characteristics, and add classification marks to the voiceprint database. At the same time, the voiceprint characteristics in the complaint information are collected and compared with the database, the relationship between customer information and voiceprint characteristics is established, malicious complaints are filtered, and customer sentiment and risk assessment are analyzed.
It realizes more accurate customer identity identification, improves the recognition and processing efficiency of malicious complaints, provides more refined services, and solves the shortcomings in the existing technology of identity identification and malicious complaint handling.
Smart Images
Figure CN120148519A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of identity recognition technology, and in particular, to a method and device for early warning of business anomalies based on voiceprint technology. Background Art
[0002] Voiceprint technology is a technology that uses voice signals for personal identity recognition. It extracts features such as the spectrum and formants of an individual's voice, establishes a voiceprint model, and compares it with an existing voiceprint database to achieve the purpose of personal identity authentication. Compared with traditional identity authentication methods, voiceprint technology has unique advantages, such as being difficult to forge and not easily stolen. Voiceprint recognition technology is a voiceprint recognition technology that identifies an individual's identity by analyzing and comparing the individual's voice signal. Compared with other voiceprint recognition technologies, voiceprint recognition technology has characteristics such as wide applicability, no need for contact, and non-invasiveness. Existing voiceprint recognition technologies mainly include three main steps: feature extraction, model training, and comparison. In the feature extraction stage, the voice signal is first preprocessed, such as removing noise and segmenting the speech; then features are extracted from the voice signal, and common features include Mel Frequency Cepstral Coefficients (MFCC), etc. In the model training stage, a large number of voiceprint samples are used for model training and optimization, and the main models used include Gaussian Mixture Model (GMM), Support Vector Machine (SVM), and deep learning models, etc. In the comparison stage, the similarity between the voiceprint features to be recognized and the known voiceprint model is calculated or matched to determine whether the individual identity matches.
[0003] Currently, in the insurance industry, in the links of complaint applications, insurance claims, and surrender of insurance policies, there are problems such as the inability to accurately identify the identity of customers, and low recognition and processing efficiency for malicious complaints. Therefore, a technical solution that applies voiceprint technology to the insurance industry is needed. Summary of the Invention
[0004] This application proposes a method and device for early warning of business anomalies based on voiceprint technology, which solves the problems in the prior art that the identity of customers cannot be accurately identified, and the recognition and processing efficiency for malicious complaints are low.
[0005] In a first aspect, an embodiment of this application provides a method for early warning of business anomalies based on voiceprint technology, including the steps of:
[0006] Collect voiceprint features and corresponding customer information, and determine whether there are preset sensitive words in the voiceprint features according to a preset rule or algorithm;
[0007] In response to the first voiceprint feature having a preset sensitive word, add a classification mark to the first voiceprint feature, and save the first voiceprint feature and the corresponding classification mark in the voiceprint database.
[0008] Further, it also includes the steps of:
[0009] Compare the voiceprint features in the collected complaint information with the voiceprint features stored in the voiceprint database;
[0010] In response to the non-existence of the second voiceprint feature in the complaint information in the voiceprint database, store the second voiceprint feature in the voiceprint database and establish information associating the customer information with the second voiceprint feature.
[0011] Further preferably, it further includes the step of:
[0012] Perform natural language processing on the complaint information to filter malicious complaints.
[0013] In one embodiment, it further includes the step of:
[0014] Collect the voiceprint features in the insurance transaction information, determine the corresponding relationship between the voiceprint features and the preset customer emotions according to the preset rules or algorithms, and output the analysis results representing the customer emotion features.
[0015] Furthermore, it further includes the step of:
[0016] Determine the corresponding relationship between the analysis results of the customer's emotions and the customer risk assessment value and insurance pricing according to the preset rules or algorithms, and output the customer risk assessment and insurance pricing data.
[0017] In one embodiment, it further includes the step of:
[0018] Collect the voiceprint features of the customer in the complaint information, compare them with the voiceprint features stored in the voiceprint database, and send an alarm message in response to the existence of the first voiceprint feature in the complaint information in the voiceprint database and the existence of a classification mark for the first voiceprint feature.
[0019] In one embodiment, it further includes the step of:
[0020] Real-time collect the voiceprint features and the corresponding customer information in any customer service program;
[0021] Compare the collected customer information with the customer information stored in the voiceprint database;
[0022] In response to the existence of the customer information in the voiceprint database and the difference between the voiceprint features corresponding to the customer information and those stored in the voiceprint database, send an alarm message.
[0023] Furthermore, it further includes the step of:
[0024] Determine the corresponding relationship between the analysis results of the customer's emotions and the customer risk assessment value and insurance pricing according to the preset rules or algorithms, and output the customer risk assessment and insurance pricing data.
[0025] Second aspect, an embodiment of the present application further provides a business anomaly warning device based on voiceprint technology, which is used to implement the business anomaly warning method based on voiceprint technology described in any one of the embodiments of the first aspect, including: an acquisition module, configured to collect voiceprint features and corresponding customer information. A determination module, configured to determine whether there is a preset sensitive word in the voiceprint features according to a preset rule or algorithm; in response to the existence of a preset sensitive word in the first voiceprint feature, adding a classification mark to the first voiceprint feature, and saving the first voiceprint feature and the corresponding classification mark in the voiceprint database.
[0026] Third aspect, an embodiment of the present application further provides a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, it implements the method described in any one of the embodiments of the first aspect.
[0027] Fourth aspect, an embodiment of the present application further provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, it implements the method described in any one of the embodiments of the first aspect.
[0028] The above at least one technical solution adopted in the embodiments of the present application can achieve the following beneficial effects:
[0029] The present application applies voiceprint technology to the insurance industry. In the links of complaint applications, insurance claims, and surrender of insurance policies, through voiceprint technology, the identity of customers can be more accurately identified, and the recognition and processing efficiency of malicious complaints can be improved. At the same time, combined with natural language processing technology, the complaint content can be classified and analyzed to provide more refined services for insurance companies. The voiceprint recognition technical solution aims to solve the limitations of the existing complaint application processing methods and improve the accuracy and efficiency of processing. Description of the Drawings
[0030] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation of the present application. In the drawings:
[0031] Figure 1 It is a system diagram of the application scenario of the present application;
[0032] Figure 2 It is a flowchart of the insurance business handling anomaly warning method provided by the embodiment of the present application;
[0033] Figure 3 It is a flowchart of the insurance business handling anomaly warning method including sensitive word judgment provided by the embodiment of the present application;
[0034] Figure 4 It is a flowchart of the warning method for the existence of classification marks in the complaint information provided by the embodiment of the present application;
[0035] Figure 5 It is a schematic diagram of the abnormal warning process for insurance business handling provided by the embodiment of the present application;
[0036] Figure 6 It is a structural diagram of the abnormal warning device for insurance business handling provided by the embodiment of the present application;
[0037] Figure 7 It is a schematic structural diagram of an electronic device provided by the embodiment of the present application. Specific embodiments
[0038] To make the purpose, technical solutions and advantages of the present application clearer, the technical solutions of the present application will be clearly and completely described below in conjunction with the specific embodiments of the present application and the corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.
[0039] The following will detail the technical solutions provided by each embodiment of the present application in conjunction with the drawings.
[0040] Figure 1 It is a system diagram of the application scenario of the present application, including terminals 11 / 12, a server 2, and a storage device 3 (storing a voiceprint database). The server and the terminals (fixed network terminals, mobile terminals) are combined into an insurance business processing system.
[0041] The terminal includes a first terminal 11 and a second terminal 12. The first terminal initiates business processes such as complaints, claims, insurance applications, and policy cancellations, and sends insurance business information and customer voice information to the server.
[0042] The storage device stores a voiceprint database. Preferably, it is a classification marker database, which can be used to train an intelligent voiceprint recognition system to determine malicious complaints and non-malicious complaints.
[0043] The voiceprint database contains a first voiceprint feature, which is the collected historical customer voiceprint data. The first voiceprint feature includes a classification marker, and the classification marker includes the presence of sensitive words and the absence of sensitive words.
[0044] The second voiceprint feature does not exist in the voiceprint database. The second voiceprint feature is a voiceprint feature that has not been collected. As the second voiceprint feature is collected, the second voiceprint feature is supplemented into the voiceprint database. Along with the second voiceprint feature supplemented into the voiceprint database, there are also the corresponding customer information of the second voiceprint feature and the corresponding classification marker of whether it contains sensitive words.
[0045] The second terminal is a customer service terminal, which receives malicious complaint alarm information and processes real complaints, etc.
[0046] The server implements voiceprint processing, malicious complaint recognition, insurance business pricing processing, and insurance business risk assessment processing;
[0047] The server and terminal system of this application loads a computer program product of the method of this application. The computer program product includes a computer program or instruction, and when the computer program or instruction is executed by a processor, it implements the method described in any embodiment of this application.
[0048] Figure 2 It is a flowchart of a service anomaly warning method based on voiceprint technology provided by an embodiment of this application, including steps 110 to 120.
[0049] Step 110: Collect voiceprint features and corresponding customer information, and judge whether there are preset sensitive words in the voiceprint features according to preset rules or algorithms.
[0050] For example, collect the first voiceprint feature and corresponding customer information, and judge whether there are preset sensitive words in the first voiceprint feature according to preset rules or algorithms;
[0051] Voice sampling and voiceprint feature extraction: When processing a complaint application, voice sample of the customer is obtained by voice sampling. Then, based on the voice sample, voiceprint features of the customer are extracted, such as spectrum, formant, etc.
[0052] Since the identity of the customer will be asked during the process of complaint application or other forms of calls with the customer, the corresponding customer information can be obtained while collecting the first voiceprint feature.
[0053] Step 120: In response to the first voiceprint feature having a preset sensitive word, add a classification mark to the first voiceprint feature.
[0054] Obtain the first voiceprint feature, and judge whether there is a preset sensitive word in the first voiceprint feature. Judge to add a classification mark to the first voiceprint feature with a preset sensitive word.
[0055] The preset sensitive word can be a word related to complaint, surrender, or claim settlement.
[0056] For example, through machine learning, a sensitive word recognition model is constructed. By inputting the customer voice sample into the sensitive word recognition model, it can be judged whether there are corresponding sensitive words in the customer's voice.
[0057] The updated voiceprint database, where the voiceprint features determined to have the sensitive word are marked.
[0058] Accurately and stably extracting the first voiceprint feature in the customer's voice is the key to achieving accurate identification. This technology requires sampling the customer's voice and extracting the distinguishable voiceprint features from it.
[0059] For example, preset relevant words for malicious complaints as sensitive words, determine them as the first voiceprint features of malicious complaints, and add them to the blacklist.
[0060] It should be noted that steps 110 to 120 are applicable to both the historical data of customer voice sampling and the real-time data of customer voice sampling. The processing of historical data generates the blacklist in the voiceprint database.
[0061] It also needs to be noted that the voiceprint database and the blacklist are two concepts. The voiceprint database contains customer information and corresponding voiceprint features, and some of the voiceprint features have classification marks, and some marks indicate malicious complaints. And the voiceprint features with malicious complaint marks constitute the blacklist.
[0062] In one of the embodiments, it further includes the steps:
[0063] In response to the existence of a classification mark for the first voiceprint feature, send an alarm message.
[0064] For example, when receiving a complaint message, if it is found that the voiceprint feature of the complaint message is the first voiceprint feature that has been added to the blacklist, send an alarm message.
[0065] The alarm message can be displayed on the monitor of the customer service staff through the GUI, or an alarm message can be sent by the way of the indicator light emitting, reminding the customer service staff that there is a risk in the current business.
[0066] In one of the embodiments, it further includes the steps:
[0067] Compare the customer information collected with the customer information stored in the voiceprint database to determine whether the customer information exists in the voiceprint database;
[0068] It should be noted that the voiceprint database stores the historical samples of customer voices. Therefore, the voiceprint database contains both voiceprint features and corresponding user information.
[0069] By comparing the customer information collected in this sampling with the customer information in the voiceprint database, it can be known whether the customer in this sampling already exists in the voiceprint database. After determining that the customer information collected in this sampling exists in the voiceprint database, by comparing the historical samples of the customer voice and the voiceprint features of this sampling, it can be determined whether the customer in this sampling is the same person.
[0070] Send an alarm message in response to the existence of the customer information in the voiceprint database and the difference between the first voiceprint feature and the voiceprint feature corresponding to the customer information stored in the voiceprint database.
[0071] Figure 3 This is a flowchart of an abnormal warning method for insurance business handling including sensitive word judgment provided by an embodiment of the present application.
[0072] Further, before determining whether there is a preset sensitive word, it further includes the steps:
[0073] Step 210: Compare the voiceprint feature in the complaint information with the voiceprint feature stored in the voiceprint database.
[0074] For example, collect the second voiceprint feature and compare it with the first voiceprint feature to determine whether the second voiceprint feature exists in the voiceprint database;
[0075] Among them, storing the first voiceprint feature in the voiceprint feature database specifically includes the steps:
[0076] Obtain feature data, input it into the first processing module, and output a voiceprint database with the first voiceprint feature;
[0077] The feature data includes the customer voice historical sampling (i.e., the voiceprint feature stored in the voiceprint database), which includes the first voiceprint feature;
[0078] The first processing module is used to determine the corresponding first voiceprint feature of the customer voice historical sampling according to a preset rule or algorithm and store it in the voiceprint database.
[0079] For example, by integrating the historical data of the customer voice in the customer voice historical sampling, collecting voiceprint data for the voice sampling of different customers, and constructing a voiceprint database.
[0080] For example, the voiceprint data of the customer can be collected through telephone recordings or an online complaint system and stored in association with the complaint information.
[0081] In the subsequent process of handling the insurance business, through the association between the voiceprint database and the customer information, by comparing the voiceprint feature of the customer with the data in the voiceprint database, the identity of the customer is determined.
[0082] For example, a voiceprint recognition engine is integrated in the system to identify and verify the voiceprint data of the customer to ensure the accuracy and authenticity of the complaint information.
[0083] Obtain the first data set, input the voiceprint database with the first voiceprint feature and the first data set into the second processing module, and output the second data set;
[0084] The first data set contains the currently sampled customer voice (i.e., the voiceprint feature in the complaint information), which includes the second voiceprint feature.
[0085] It should be noted that in this application, the first voiceprint feature refers to the voiceprint feature collected historically, while the second voiceprint feature refers to the currently collected voiceprint feature.
[0086] For example, when a customer contacts an insurance company by phone to discuss matters related to complaint applications, insurance claims, or policy cancellations, etc., the currently sampled customer voice obtained by answering the phone is used to extract the voiceprint information in the currently sampled customer voice.
[0087] The second processing module is used to compare the second voiceprint feature of the currently sampled customer voice with the first voiceprint feature in the voiceprint database to find the first voiceprint feature that is consistent with the second voiceprint feature.
[0088] The second voiceprint feature extracted from the currently sampled customer voice is compared with the first voiceprint feature in the voiceprint database to determine whether there is a first voiceprint feature in the voiceprint database that is consistent with the second voiceprint feature in the currently sampled customer voice. If there is a first voiceprint feature that is consistent with the second voiceprint feature, it is considered that the second voiceprint feature belongs to a customer existing in the voiceprint database.
[0089] The second data set contains the comparison result of whether the voiceprint of the currently sampled customer voice exists in the voiceprint database.
[0090] The comparison result includes that the second voiceprint feature exists in the voiceprint database and that the second voiceprint feature does not exist in the voiceprint database.
[0091] Step 220: In response to the non-existence of the second voiceprint feature in the complaint information in the voiceprint database, store the second voiceprint feature in the voiceprint database and establish information associating the customer information with the second voiceprint feature.
[0092] The non-existence of the second voiceprint feature in the voiceprint database indicates that this customer is not an existing customer.
[0093] By establishing a malicious complaint model, known malicious complaints are marked and classified, and matched with the first voiceprint feature to distinguish malicious complaints from genuine complaints, improving the processing efficiency and accuracy.
[0094] For example, in response to the non-existence of the second voiceprint feature in the voiceprint database, store the second voiceprint feature in the voiceprint database and establish information associating the customer information with the second voiceprint feature.
[0095] A new entry, namely the voiceprint feature and the corresponding customer information, is established in the voiceprint database for the second voiceprint feature that does not exist in the voiceprint database.
[0096] Compare the second voiceprint feature of the current customer voice sample. If a second voiceprint feature that does not exist in the voiceprint database is found, it can be entered into the voiceprint database as a new customer voiceprint feature.
[0097] If the second voiceprint feature in the current customer voice sample exists in the voiceprint database, then it is considered that the second voiceprint feature belongs to an existing customer.
[0098] Further preferably, in step 230, perform natural language processing on the complaint information to filter malicious complaints.
[0099] In step 240, similar to 110, collect the voiceprint feature and the corresponding customer information, and determine whether there is a preset sensitive word in the voiceprint feature according to a preset rule or algorithm.
[0100] In step 250, similar to 120, in response to the first voiceprint feature having a preset sensitive word, add a classification mark to the first voiceprint feature, and save the first voiceprint feature and the corresponding classification mark in the voiceprint database.
[0101] In one embodiment, as Figure 4 shown, before determining whether there is a preset sensitive word, the steps further include:
[0102] In step 310, similar to 110, collect the voiceprint feature and the corresponding customer information, and determine whether there is a preset sensitive word in the voiceprint feature according to a preset rule or algorithm.
[0103] In step 320, similar to 120, in response to the first voiceprint feature having a preset sensitive word, add a classification mark to the first voiceprint feature, and save the first voiceprint feature and the corresponding classification mark in the voiceprint database.
[0104] In step 330, collect the voiceprint feature of the customer in the complaint information, compare it with the voiceprint feature stored in the voiceprint database, and in response to the first voiceprint feature in the complaint information existing in the voiceprint database and the first voiceprint feature having a classification mark, send an alarm message.
[0105] In response to the first voiceprint feature existing in the voiceprint database having a classification mark, send an alarm message.
[0106] Obtain the customer voice sample, obtain the second voiceprint feature of the customer voice sample, compare it with the voiceprint database, and if the second voiceprint feature exists in the voiceprint database and the second voiceprint feature has a classification mark, directly generate classification information and give a reminder to the customer service.
[0107] It should be noted that in the above two embodiments, in one embodiment, the customer voice is sampled, and the sampled second voiceprint feature is used to identify preset sensitive words. If sensitive words are found, a classification information reminder is sent to the customer service. At the same time, a classification mark is added to the voiceprint feature and it is added to the blacklist.
[0108] In another embodiment, the customer voice is sampled, and it is found that the sampled second voiceprint feature exists in the voiceprint database and has a mark, which indicates that this voiceprint feature has a historical record of uttering preset sensitive words and has been marked. At this time, regardless of whether this first voiceprint feature is recognized as uttering preset sensitive words this time, a classification information reminder is sent to the customer service.
[0109] For example, through machine learning, a sensitive word recognition processing module is constructed. By inputting the sampled customer voice into the sensitive word recognition processing module, it can be determined whether there are corresponding sensitive words in the customer's voice.
[0110] In response to the existence of a preset sensitive word in the first voiceprint feature, a mark is added to the first voiceprint feature in the voiceprint database, and classification information is generated.
[0111] In the process of sampling the customer voice to construct the voiceprint database, through the recognition of sensitive words in the customer voice, a mark is added to the first voiceprint feature that recognizes sensitive words, and it is added to the voiceprint blacklist.
[0112] For example, to add a mark to the first voiceprint feature of the preset sensitive word and add it to the voiceprint blacklist, classify and record that this voiceprint feature enters the blacklist, and generate classification information.
[0113] The classification information can be a reminder to the customer service that conducts a voice call with the customer from the display interface or voice prompt aspect, reminding the customer service to perform abnormal processing.
[0114] In one of the embodiments, it further includes the steps of:
[0115] Collect the voiceprint features in the insurance transaction information, determine the corresponding relationship between the voiceprint features and the preset customer emotions according to the preset rules or algorithms, and output the analysis result representing the customer emotion characteristics;
[0116] Determine the corresponding relationship between the second voiceprint feature and the preset customer emotion according to the preset rules or algorithms, and output the analysis result of the customer emotion.
[0117] Obtain the voiceprint database with the first voiceprint feature, input it into the fourth processing module, and output the fifth data set;
[0118] The fourth processing module is used to determine the corresponding relationship between the voiceprint database or the first data set with the first voiceprint feature and the preset customer emotion.
[0119] The fifth data set contains the analysis results of the emotions of customers.
[0120] For example, voiceprint recognition technology can be combined with technologies such as natural language processing to intelligently analyze and recognize emotions in customer complaint information, helping insurance companies better understand customer needs and solve problems.
[0121] Insurance companies can use voiceprint recognition technology to monitor and evaluate the call quality of telephone salespersons, improving sales efficiency and service quality. At the same time, they can also analyze the emotions and needs of customers through voiceprint recognition technology to achieve more intelligent customer communication.
[0122] Furthermore, it also includes the steps of:
[0123] Determine the corresponding relationship between the analysis results of the emotions of customers, customer risk assessment values, and insurance pricing according to preset rules or algorithms, and output customer risk assessment and insurance pricing data.
[0124] Input the fifth data set into the fifth processing module to output a sixth data set; the fifth processing module is used to determine the corresponding relationship between the analysis results of the emotions of customers, customer risk assessment values, and insurance pricing according to preset rules or algorithms.
[0125] The sixth data set is customer risk assessment and insurance pricing data.
[0126] Voiceprint recognition technology can be combined with other data sources, such as voice emotion analysis, speech rate, etc., for customer risk assessment and insurance pricing, improving the risk management ability of insurance companies.
[0127] In one embodiment, it also includes the step of: real-time collecting voiceprint features and corresponding customer information in any customer service program.
[0128] For example, as Figure 5 shown, when a customer service agent answers a customer call, voiceprint feature comparison is performed during the answering process. If the corresponding voiceprint feature is found in the voiceprint database, the customer identity can be directly confirmed. If the voiceprint feature is not found, it means this is a new customer, and then the current sampling of the customer's voice is extracted to collect new voiceprint features.
[0129] In one embodiment, it also includes the step of: comparing the collected customer information with the customer information stored in the voiceprint database.
[0130] In response to the existence of the customer information in the voiceprint database and the voiceprint feature corresponding to the customer information being different from that stored in the voiceprint database, send an alarm message.
[0131] For the identified customers, determine whether they have a history of uttering the set sensitive words by checking whether the corresponding voiceprint features in the voiceprint database are marked. If so, directly send the classification information to the customer service.
[0132] If there is no history of uttering the set sensitive words, or the customer is a new customer, then identify the voiceprint features of this call to determine whether there are set sensitive words. If there are set sensitive words, mark the voiceprint features and send the classification information to the customer service.
[0133] If there are no set sensitive words, no classification information is sent.
[0134] Figure 6 This application embodiment also provides a structural diagram of a service anomaly warning device based on voiceprint technology, which is used to implement the service anomaly warning method based on voiceprint technology described in any embodiment of the first aspect, and includes:
[0135] An acquisition module 410, which is used to collect voiceprint features and corresponding customer information;
[0136] A determination module 420, which is used to determine whether there are preset sensitive words in the voiceprint features according to preset rules or algorithms; and is also used to add a classification mark to the first voiceprint feature in response to the existence of preset sensitive words in the first voiceprint feature, and save the first voiceprint feature and the corresponding classification mark in the voiceprint database.
[0137] In one embodiment, the acquisition module includes a first acquisition unit, which is used to collect voiceprint features and corresponding customer information.
[0138] The determination module includes a first determination unit, which is used to determine whether there are preset sensitive words in the voiceprint features according to preset rules or algorithms. It also includes a second determination unit, which is used to add a classification mark to the first voiceprint feature in response to the existence of preset sensitive words in the first voiceprint feature, and save the first voiceprint feature and the corresponding classification mark in the voiceprint database.
[0139] In one embodiment, the determination module further includes a third determination unit, which is used to compare the first voiceprint feature with the voiceprint features stored in the voiceprint database to determine whether the first voiceprint feature exists in the voiceprint database.
[0140] It further includes a fourth determination unit, which is used to store the first voiceprint feature in the voiceprint database and establish the information associating the customer information and the first voiceprint feature in response to the non-existence of the first voiceprint feature in the voiceprint database.
[0141] The above embodiments are used to implement any one of the steps 210 to 250.
[0142] In one embodiment, an output module 430 is further included. The output module includes a first output unit configured to determine the correspondence between the first voiceprint feature and the preset customer emotion according to a preset rule or algorithm, and output the analysis result of the customer emotion.
[0143] The above embodiment is used to implement any one of the first voiceprint feature and the preset customer emotion in the embodiment.
[0144] In one embodiment, an output module is further included. The output module includes a second output unit configured to send an alarm message in response to the presence of a classification mark in the first voiceprint feature.
[0145] The above embodiment is used to implement any one of the embodiments where a classification mark exists in the first voiceprint feature in step 120.
[0146] In one embodiment, the determination module further includes a fifth determination unit configured to compare the collected customer information with the customer information stored in the voiceprint database to determine whether the customer information exists in the voiceprint database.
[0147] An output module is further included. The output module includes a third output unit configured to send an alarm message in response to the existence of the customer information in the voiceprint database and the difference between the first voiceprint feature and the voiceprint feature corresponding to the customer information stored in the voiceprint database.
[0148] The above embodiment is used to implement any one of the embodiments where the customer information stored in the voiceprint database is compared in step 120.
[0149] In one embodiment, an output module is further included. The output module includes a fourth output unit configured to send an alarm message in response to the first voiceprint feature existing in the voiceprint database with a classification mark.
[0150] In one embodiment, an output module is further included. The output module includes a fifth output unit configured to determine the correspondence between the analysis result of the customer's emotion and the customer risk assessment value and insurance pricing according to a preset rule or algorithm, and output the customer risk assessment and insurance pricing data.
[0151] The above embodiment is used to implement any one of the embodiments of the correspondence between the analysis result of the customer's emotion and the customer risk assessment value and insurance pricing.
[0152] The above embodiment is used to implement any one of calculating the customer risk assessment and insurance pricing based on the analysis result of the customer's emotion and demand.
[0153] To implement the methods of the embodiments of the present application, such as Figure 1As shown in the figure, the embodiment of the present application further provides a business anomaly warning system based on voiceprint technology, including: a computer device, configured to run the business anomaly warning method based on voiceprint technology according to any one of the embodiments of the first aspect, and generate data with classification marks added to the first voiceprint feature. For example, the computer device loads Figure 6 each software operation module of the embodiment shown.
[0154] In one embodiment, it further includes a storage device 3.
[0155] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
[0156] Therefore, the present application also proposes a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, it implements the method described in any one of the embodiments of the present application.
[0157] The present invention is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present invention. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, as well as the combination of flows and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for implementing the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.
[0158] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured product including an instruction device, and the instruction device implements the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.
[0159] These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process. Thus, the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one process Figure 1 one process or multiple processes and / or blocks Figure 1 steps for implementing the functions specified in one block or multiple blocks.
[0160] Furthermore, the present application also provides an electronic device (or computing device), including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, the method described in any embodiment of the present application is implemented.
[0161] In a typical configuration, a computing device includes one or more processors (CPUs), an input / output interface, a network interface, and a memory. The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM), and / or non-volatile memory in the form of, for example, read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.
[0162] Figure 7 It is a schematic structural diagram of an electronic device provided by an embodiment of the present application. The displayed electronic device 600 is only an example and should not impose any limitation on the functions and usage scope of the embodiments of the present application. It includes: one or more processors 620; a storage device 610 for storing one or more programs. When the one or more programs are run by the one or more processors 620, the one or more processors 620 implement the business anomaly warning method based on voiceprint technology provided by the embodiments of the present application. The method includes: collecting voiceprint features and corresponding customer information, and judging whether there are preset sensitive words in the voiceprint features according to a preset rule or algorithm; in response to the first voiceprint feature having a preset sensitive word, adding a classification mark to the first voiceprint feature, and saving the first voiceprint feature and the corresponding classification mark in a voiceprint database. Preferably, the method further includes any step of the method in the foregoing embodiments of the present application.
[0163] The electronic device 600 further includes an input device 630 and an output device 640; the processor 620, the storage device 610, the input device 630, and the output device 640 in the electronic device can be connected by a bus or other means. In the figure, it is taken as an example of being connected by a bus 650.
[0164] The storage device 610, as a computer-readable storage medium, can be used to store software programs, computer-executable programs, and module units, such as the program instructions corresponding to the method for determining cloud base height in the embodiments of the present application. The storage device 610 mainly includes a program storage area and a data storage area. Among them, the program storage area can store an operating system and application programs required for at least one function; the data storage area can store data created according to the use of the terminal and the like. In addition, the storage device 610 can include high-speed random access memory, and can also include non-volatile memory, such as at least one magnetic disk storage device, a flash memory device, or other non-volatile solid-state storage devices. In some instances, the storage device 610 can further include a memory remotely set relative to the processor 620, and these remote memories can be connected through a network. Examples of the above network include but are not limited to the Internet, an enterprise intranet, a local area network, a mobile communication network, and combinations thereof.
[0165] The input device 630 can be used to receive input digital, character information, or voice information, and generate key signal inputs related to the user settings and function controls of the electronic device. The output device 640 can include electronic devices such as a display screen and a speaker.
[0166] It should also be noted that the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or further includes elements inherent to such process, method, commodity or device. Without further limitation, an element defined by the statement "including one..." does not exclude the existence of another identical element in the process, method, commodity or device including the said element.
[0167] The above are only the embodiments of the present application and are not used to limit the present application. For those skilled in the art, the present application can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included within the scope of the claims of the present application.
Claims
1. A business abnormality early warning method based on voiceprint technology, characterized in that: Includes steps: Collect voiceprint features and corresponding customer information, and determine whether there are preset sensitive words in the voiceprint features based on preset rules or algorithms; In response to the presence of a preset sensitive word in the first voiceprint feature, a classification mark is added to the first voiceprint feature, and the first voiceprint feature and the corresponding classification mark are saved in a voiceprint database.
2. According to claim 1, the service abnormality early warning method based on voiceprint technology is characterized in that: Also includes the steps: Collect the voiceprint features in the complaint information and compare them with the voiceprint features stored in the voiceprint database; In response to the second voiceprint feature in the complaint information not existing in the voiceprint database, the second voiceprint feature is stored in the voiceprint database and information associating the customer information with the second voiceprint feature is established.
3. According to claim 1, the service abnormality early warning method based on voiceprint technology is characterized in that: Also includes the steps: Collect voiceprint features in insurance transaction information, determine the correspondence between the voiceprint features and preset customer emotions according to preset rules or algorithms, and output analysis results representing customer emotional characteristics.
4. According to claim 1, the service abnormality early warning method based on voiceprint technology is characterized in that: Also includes the steps: The voiceprint features of the customer are collected in the complaint information and compared with the voiceprint features stored in the voiceprint database. In response to the presence of the first voiceprint feature in the complaint information in the voiceprint database and the presence of a classification mark in the first voiceprint feature, an alarm message is sent.
5. According to claim 1, the method for early warning of business anomalies based on voiceprint technology is characterized in that: Also includes the steps: Collect voiceprint features and corresponding customer information in real time in any customer service program; Compare the collected customer information with the customer information stored in the voiceprint database; In response to the customer information existing in the voiceprint database, and the voiceprint feature corresponding to the customer information is different from that stored in the voiceprint database, an alarm message is sent.
6. The service abnormality early warning method based on voiceprint technology according to claim 2 is characterized in that: Also includes the steps: Perform natural language processing on complaint information to filter out malicious complaints.
7. The method for early warning of abnormal business operations based on voiceprint technology according to claim 3 is characterized in that: Also includes the steps: The correspondence between the analysis results of the customer's emotions and the customer's risk assessment value and insurance pricing is determined according to preset rules or algorithms, and the customer's risk assessment and insurance pricing data are output.
8. A business abnormality warning device based on voiceprint technology, characterized in that: The method for implementing the business abnormality warning method based on voiceprint technology as described in any one of claims 1 to 7 comprises: Acquisition module, used to collect voiceprint features and corresponding customer information; The determination module is used to determine whether there are preset sensitive words in the voiceprint feature according to preset rules or algorithms; it is also used to add a classification mark to the first voiceprint feature in response to the presence of preset sensitive words in the first voiceprint feature, and save the first voiceprint feature and the corresponding classification mark in the voiceprint database.
9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.
10. An electronic device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the method according to any one of claims 1 to 7 is implemented.