Retrieval method, device and equipment for insurance cases and storage medium

Abstract

The invention relates to the technical field of information retrieval, and provides an insurance case retrieval method and device, equipment and a storage medium, and the method comprises the steps: obtaining a text input by a user; obtaining a to-be-retrieved verbal skill text corresponding to the text input by the user; determining a target verbal skill vector corresponding to the to-be-retrieved verbal skill text from preset verbal skill vectors included in a verbal skill vector library according to the to-be-retrieved verbal skill text; according to the target verbal skill vector, determining retrieval prompt information corresponding to the verbal skill text to be retrieved; and according to the retrieval prompt information, determining a target insurance case corresponding to the verbal skill text to be retrieved from a plurality of preset insurance cases, so as to improve the retrieval convenience of the insurance cases.

Description

Technical Field

[0001] The present application relates to the field of information retrieval technology, and in particular to a retrieval method, device, equipment and storage medium for insurance reporting cases. Background Art

[0002] At present, with the development of insurance business in the financial industry, fraudulent insurance fraud is frequent. Insurance fraudsters usually commit crimes in teams. For example, insurance fraudsters usually use different phone numbers or the assistance of others to make multiple insurance reports according to similar insurance reporting scripts to achieve the purpose of insurance fraud. Based on this, insurance business personnel need to identify the insurance reporting scripts of each insurance case to determine whether the insurance reporting script is an insurance fraud script.

[0003] However, in the case of a large number of insurance reports, if business personnel need to verify whether there are any insurance reports that use similar insurance report scripts among the large number of insurance reports, it will take a lot of manpower and time to identify the insurance report scripts of each insurance report, resulting in poor convenience in retrieving insurance reports. Summary of the invention

[0004] The main purpose of this application is to provide a retrieval method, device, equipment and storage medium for insurance cases, aiming to improve the convenience of retrieval of insurance cases.

[0005] In a first aspect, the present application provides a method for retrieving insurance cases, comprising the following steps:

[0006] Get user input text;

[0007] Obtain the to-be-searched speech text corresponding to the user input text;

[0008] According to the to-be-searched speech text, determine the target speech vector corresponding to the to-be-searched speech text from the preset speech vectors included in the speech vector library;

[0009] Determine, according to the target speech vector, the search prompt information corresponding to the speech text to be searched;

[0010] According to the search prompt information, a target insurance case corresponding to the to-be-searched speech text is determined from a plurality of preset insurance cases.

[0011] In a second aspect, the present application further provides a device for searching insurance cases, the device comprising:

[0012] A first text acquisition module, used to acquire user input text;

[0013] A second text acquisition and determination module is used to acquire the to-be-searched speech text corresponding to the user input text;

[0014] A vector determination module, used to determine, according to the to-be-retrieved speech text, a target speech vector corresponding to the to-be-retrieved speech text from preset speech vectors included in a speech vector library;

[0015] A search prompt determination module, used to determine the search prompt information corresponding to the to-be-searched speech text according to the target speech vector;

[0016] The case determination module is used to determine the target insurance case corresponding to the to-be-searched speech text from a plurality of preset insurance cases according to the search prompt information.

[0017] In a third aspect, the present application further provides a computer device, the computer device comprising a memory and a processor;

[0018] The memory is used to store computer programs;

[0019] The processor is used to execute the computer program and implement the steps of the above-mentioned insurance case retrieval method when executing the computer program.

[0020] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of the above-mentioned insurance case retrieval method are implemented.

[0021] The present application provides a method, device, equipment and storage medium for searching insurance cases. The method for searching insurance cases includes: obtaining a user input text; obtaining a text to be searched corresponding to the user input text; determining a target text vector corresponding to the text to be searched from a preset text vector included in a text vector library according to the text to be searched; determining search prompt information corresponding to the text to be searched according to the target text vector; determining a target insurance case corresponding to the text to be searched from multiple preset insurance cases according to the search prompt information, so as to improve the convenience of searching insurance cases. For example, it can be used for insurance business in the financial field. When handling insurance business, if a business person receives a call from an insurance person, the user input text can be determined according to the language description of the insurance person, and the target insurance case related to the text to be searched corresponding to the user input text can be obtained according to the search method of the insurance case. The business person can refer to the target insurance case to determine whether the insurance person has committed insurance fraud, such as whether the language description of the insurance person involves insurance fraud speech. Accordingly, if the business personnel need to search for different target insurance cases corresponding to the same insurance reporting script, the business personnel can input the relevant user input text, and through the insurance case search method, obtain the target insurance cases related to the to-be-searched script text corresponding to the user input text. Based on this, in the process of searching for target insurance cases, there is no need to spend a lot of manpower and time to manually check multiple preset insurance cases, which is conducive to improving the convenience of searching for insurance cases, thereby helping to improve the processing efficiency of insurance claims business. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0023] Figure 1 It is a flowchart of a method for retrieving insurance cases provided in an embodiment of the present application;

[0024] Figure 2 It is a schematic diagram of the search process of insurance reporting cases involved in an embodiment of the present application;

[0025] Figure 3 It is a schematic diagram of a process for determining a speech vector library involved in an embodiment of the present application;

[0026] Figure 4 It is a schematic block diagram of a retrieval device for insurance cases provided in an embodiment of the present application;

[0027] Figure 5It is a schematic block diagram of the structure of a computer device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0028] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0029] The flowcharts shown in the accompanying drawings are only examples and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may also be decomposed, combined or partially merged, so the actual execution order may change according to actual conditions.

[0030] The embodiments of the present application provide a method, device, equipment and storage medium for retrieving insurance cases. The method for retrieving insurance cases can be applied to a computer device, which can be a mobile phone, a tablet computer, a laptop computer, a desktop computer and the like. The method can also be applied to a server, which can be a separate server or 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 (CDNs), and big data and artificial intelligence platforms.

[0031] In conjunction with the accompanying drawings, some embodiments of the present application are described in detail below. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other.

[0032] See also Figure 1 , Figure 1 It is a flowchart of a method for retrieving insurance cases provided in an embodiment of the present application. It should be noted that the method for retrieving insurance cases provided in an embodiment of the present application can be used in a computer device, and of course can also be used in a server. For example, the server can obtain user input text from a computer device, process the user input text according to the method for retrieving insurance cases, and obtain a target insurance case. Exemplarily, the server can, for example, send the target insurance case to a computer device; of course, it is not limited to this, and no limitation is made here.

[0033] In specific implementation, computer devices include but are not limited to: any one of mobile phones, tablet computers, laptop computers, and desktop computers; the server can be a single server or a server cluster, or a cloud server that provides cloud computing services.

[0034] like Figure 1 As shown, the insurance case retrieval method includes steps S101 to S105.

[0035] Step S101: Obtain user input text.

[0036] For example, in the case of detecting a text input operation of a user on a computer device, the near-eye display device may obtain the user input text in response to the user's text input operation.

[0037] Take the insurance business as an example. When a business person receives a call from an insurance reporting person about a current insurance case, the business person can input text into a computer device based on the preset insurance reporting voice provided by the insurance reporting person, and the computer device can obtain the user input text. Of course, this is not limited to this. When a business person needs to retrieve an insurance case related to a corresponding insurance reporting script, he can input text into a computer device based on his memory of the insurance reporting script, and the computer device can obtain the user input text. This is not limited here.

[0038] In this way, by obtaining the user input text, the user input text can be used to subsequently determine the corresponding to-be-searched speech text, which is conducive to subsequently improving the convenience of determining the to-be-searched speech text.

[0039] Step S102: Obtain the to-be-searched speech text corresponding to the user input text.

[0040] For example, when the user input text is obtained, the user input text includes, for example, idiomatic words, logical relationship words, and other texts that have a low correlation with the insurance reporting language. In order to avoid the adverse effects of texts that have a low correlation with the insurance reporting language on the retrieval efficiency and retrieval convenience of insurance reporting cases, the computer device can perform data cleaning processing on the user input text, remove the texts that have a low correlation with the insurance reporting language in the user input text, and retain the texts that have a high correlation with the insurance reporting language in the user input text, so as to obtain the corresponding language text to be retrieved.

[0041] The to-be-searched speech text may be used to indicate the to-be-searched insurance reporting speech. The to-be-searched insurance reporting speech may include at least one of the insurance reporting speech of the insurance reporting personnel and the insurance reporting speech that the business personnel need to check, which is not limited here.

[0042] Take the insurance business as an example. When a business person receives a call from an insured person, the user input text acquired by the computer device may be the text input by the business person based on the description of the insured person. The user input text may include the colloquial description idioms of the insured person, and the colloquial description idioms have a low correlation with the insurance reporting speech. The computer device may remove the colloquial description idioms in the user input text to determine the speech text to be retrieved corresponding to the user input text.

[0043] Of course, it is not limited to this. The user input text obtained by the computer device can be text input by the business personnel based on their own memory information of the insurance reporting script. The user input text, for example, includes the business personnel's text description idioms, and the text description idioms have a low correlation with the insurance reporting script. The computer device can then remove the text description idioms in the user input text to determine the script text to be retrieved corresponding to the user input text. This is not limited here.

[0044] In this way, by obtaining the to-be-retrieved speech text corresponding to the user input text, the to-be-retrieved speech text can be used for subsequent determination of the target speech vector, which is conducive to subsequent improvement of the convenience of determining the target speech vector.

[0045] Step S103: According to the speech text to be retrieved, determine the target speech vector corresponding to the speech text to be retrieved from the preset speech vectors included in the speech vector library.

[0046] Exemplarily, the computer device may be provided with a speech vector library. The speech vector library may store a plurality of preset speech vectors. The preset speech vector may be used to indicate the vector corresponding to the insurance speech of the currently processed preset insurance case.

[0047] In the case where the text of the to-be-searched speech is obtained, the computer device may search the speech vector library based on the text of the to-be-searched speech to determine whether the speech of the currently processed preset insurance case indicated by the preset speech vector in the search speech vector library is similar to the to-be-searched insurance speech indicated by the text of the to-be-searched speech. In the case where the computer device determines that the speech vector library contains the preset speech vector indicating the currently processed preset insurance case and the to-be-searched insurance speech indicated by the text of the to-be-searched speech, the computer device may determine the preset speech vector as the target speech vector corresponding to the text of the to-be-searched speech.

[0048] In some embodiments, the speech text to be retrieved is embedded to obtain a speech vector to be retrieved corresponding to the speech text to be retrieved; the similarity between the speech vector to be retrieved and a preset speech vector in a speech vector library is determined; and the preset speech vector whose similarity is greater than or equal to a preset similarity threshold is determined as the target speech vector corresponding to the speech text to be retrieved.

[0049] like Figure 2 As shown, the computer device can embed the to-be-searched speech text to obtain the to-be-searched speech vector corresponding to the to-be-searched speech text. For example, the computer device can input the to-be-searched speech text into a preset embedding model, and the preset embedding model can embed the to-be-searched speech text and output the corresponding to-be-searched speech vector.

[0050] like Figure 2 As shown, when the speech vector to be retrieved is obtained, the computer device can make a similarity judgment based on the speech vector to be retrieved and the preset speech vector provided by the speech vector library to determine the similarity between the speech vector to be retrieved and each preset speech vector in the speech vector library. Accordingly, the computer device can select the preset speech vector with a similarity greater than or equal to the preset similarity threshold as the target speech vector based on the similarity between the speech to be retrieved and each preset speech vector.

[0051] For example, the computer device may input the to-be-retrieved speech vector into a preset language model so that the preset language model can retrieve the preset speech vector in the dialogue vector library based on the to-be-retrieved speech vector. For example, the preset language model may include a language model based on the LangChain architecture, and the preset language model can retrieve the preset speech vector in the dialogue vector library based on a retriever, such as a retriever, in combination with the to-be-retrieved speech vector. In the process of retrieving the preset speech vector in the speech vector library, the preset language model can determine the similarity between the to-be-retrieved speech vector and the preset speech vector. Accordingly, in the case where the similarity between the preset speech vector and the to-be-retrieved speech vector is greater than or equal to the preset similarity threshold, the preset language model can determine the preset speech vector as the target speech vector and output the target speech vector. Accordingly, the computer device can determine the target speech vector corresponding to the to-be-retrieved speech based on the target speech vector output by the preset language model.

[0052] In an exemplary embodiment, when there are multiple preset speech vectors in the speech vector library and the similarities between them and the speech vector to be retrieved are all greater than or equal to a preset similarity threshold, the computer device can select the preset speech vector with the highest similarity as the target speech vector corresponding to the speech text to be retrieved.

[0053] In this way, when determining the speech text to be retrieved, the computer device can combine the speech vector library to determine the target speech vector corresponding to the speech text to be retrieved, which is conducive to improving the convenience of determining the target speech vector. Correspondingly, the target speech vector can be used for subsequent determination of search prompt information, which is conducive to improving the convenience of subsequent determination of search prompt information.

[0054] Exemplarily, the computer device may be pre-built with a speech vector library.

[0055] In some embodiments, a preset insurance report voice corresponding to a preset insurance report case is obtained; a preset speech text corresponding to the preset insurance report voice is obtained; semantic segmentation is performed on the preset speech text to obtain a preset speech text segment corresponding to the preset speech text; embedding is performed on the preset speech text segment to obtain a preset speech vector corresponding to the preset speech text; and a speech vector library is determined based on the preset speech vectors corresponding to each of the multiple preset speech texts.

[0056] The preset insurance case can be used to indicate the processed insurance case. In the past, when the business personnel received the insurance call from the insurance personnel for the preset insurance case, they could record the call process of the insurance call to form the preset insurance voice. Based on this, the computer device can obtain the preset insurance voice.

[0057] like Figure 3 As shown, in the case where the computer device obtains the preset insurance reporting voice, the computer device can perform corresponding data processing on the preset insurance reporting voice to obtain the preset speech text corresponding to the preset insurance reporting voice.

[0058] In some embodiments, a preset insurance report voice is subjected to voice conversion processing to obtain a preset insurance report text corresponding to the preset insurance report voice; the preset insurance report text is subjected to data cleaning processing to obtain an initial speech text; and the preset speech text corresponding to the preset insurance report voice is determined based on the initial speech text and the speech tag corresponding to the initial speech text.

[0059] For example, the computer device can perform voice conversion processing on the preset insurance report voice through a preset voice-to-text algorithm to obtain a preset insurance report text. The voice-to-text algorithm includes, for example, an automatic speech recognition (Automatic Speech Recognition, ASR) algorithm, which is not limited here.

[0060] When the preset insurance report text is obtained, since the business personnel and the insurance reporter can communicate with each other during the call, the preset insurance report text may contain text that is irrelevant to the insurance reporter's insurance report speech. Based on this, the computer device can perform data cleaning processing on the preset insurance report text to filter out text suitable for extracting the insurance reporter's insurance report speech from the preset insurance report text to form an initial speech text.

[0061] For example, when the initial speech text is determined, the speech label corresponding to the initial speech text can be marked. The computer device can determine the initial speech text carrying the speech label as the preset speech text, so that the speech vector library can be constructed based on the preset speech text. For example, the computer device can load the initial speech text carrying the speech label through a loader, and convert the speech label and the initial speech text into the preset speech text.

[0062] In an exemplary implementation, in the case where there is a case where the insured person defrauds insurance according to the speech corresponding to the initial speech text in the preset insurance case, the business personnel can set the speech tag corresponding to the initial speech text as the insurance fraud speech, and then the computer device can combine all the initial speech texts with the speech tags of the insurance fraud speech to build a speech vector library. Accordingly, the computer device can subsequently search for the preset insurance case related to the corresponding insurance fraud speech based on the speech vector library.

[0063] The preset speech text can be used to indicate the text corresponding to the insurance speech of the currently processed preset insurance case. When the computer obtains the preset speech text corresponding to the preset insurance voice, the computer device can determine the preset speech vector according to the preset speech text.

[0064] like Figure 3 As shown, the computer device can perform semantic segmentation processing on the preset speech text through a preset semantic segmentation algorithm to obtain preset speech text segments corresponding to the preset speech text. The preset semantic segmentation algorithm includes, for example, a text splitter algorithm, which is not limited here. In an exemplary embodiment, if the preset insurance case is directly retrieved through the preset speech text, the hit rate of retrieving the preset insurance case will decrease accordingly as the text length of the preset speech text increases. Based on this, the computer device can improve the hit rate of subsequent retrieval of the preset insurance case by segmenting the preset speech text into preset speech text segments.

[0065] like Figure 3 As shown, when the computer device obtains the preset speech text segment, it can embed the preset speech text to obtain the corresponding preset speech text segment. For example, the computer device can determine the preset speech vector through a preset preset embedding model. For example, the computer device can input the preset speech text segment into the preset embedding model so that the preset embedding model embeds the preset speech text segment to obtain and output the preset speech vector corresponding to the preset speech text.

[0066] When the computer device determines the preset speech vector corresponding to the preset speech text, the computer device can construct a speech vector library by integrating the preset speech vectors corresponding to the preset speech texts corresponding to all preset insurance reporting cases.

[0067] In some implementations, a plurality of preset speech vectors are stored according to a preset data storage algorithm to obtain a speech vector library.

[0068] For example, the quantity storage algorithm may include an unstructured data storage algorithm. The computer device may store a plurality of preset speech vectors through the unstructured data storage algorithm to obtain a speech vector library.

[0069] In this way, the computer device can obtain the corresponding preset speech text through the preset insurance reporting voice corresponding to the preset insurance reporting case, and determine the preset speech text segment based on the preset speech text, thereby determining the corresponding preset speech vector based on the preset speech text segment, and combining multiple preset speech vectors to determine the speech vector library, which is conducive to improving the convenience of determining the speech vector library. Correspondingly, the speech vector library can be used to determine the target speech vector corresponding to the speech text to be retrieved, which is conducive to improving the convenience of determining the target speech vector.

[0070] Step S104: Determine the search prompt information corresponding to the to-be-searched speech text according to the target speech vector.

[0071] When the target speech vector is determined, the computer device can determine the search prompt information corresponding to the speech text to be searched according to the target speech vector. The search prompt information can be used to assist in searching the target insurance case corresponding to the speech text to be searched.

[0072] For example, since the target speech vector is selected from the preset speech vectors included in the speech vector library based on the speech text to be retrieved, the speech text to be retrieved corresponding to the target speech vector and the preset speech text corresponding to the preset speech vector may be different. For example, the speech text to be retrieved is determined based on the user input text, and the preset speech text is determined based on the preset insurance case. The user input text can be determined based on the input information of the business personnel, and the preset insurance case can be determined based on the description of the insurance personnel. In the case where the business personnel and the insurance personnel are different users, the user input text and the description involved in the preset insurance case may be different. Based on this, when searching for the preset insurance case, the computer device needs to determine the search prompt information corresponding to the speech text to be retrieved. The search prompt information can be used to establish an association relationship between the speech text to be retrieved corresponding to the target speech vector and the preset speech text corresponding to the target speech vector. Then, based on the association relationship, the computer device can search for the preset insurance case according to the search prompt information associated with the preset insurance case. The search prompt information may include a preset speech text corresponding to the target speech vector, and may also include a preset speech text segment corresponding to the preset speech text corresponding to the target speech vector.

[0073] In this way, by determining the search prompt information corresponding to the speech text to be retrieved based on the target speech vector, the search prompt information can establish a correlation between the speech text to be retrieved corresponding to the target speech vector and the preset speech text corresponding to the target speech vector. The computer device can subsequently search for the preset insurance cases based on the prompt of the search prompt information, which is conducive to improving the convenience of searching for the preset insurance cases.

[0074] In some embodiments, based on the association between a preset speech vector and a preset speech text segment, according to the target speech vector, a target speech text segment corresponding to the target speech vector is determined from the preset speech text segments; and the target speech text segment is determined as the retrieval prompt information corresponding to the speech text to be retrieved.

[0075] For example, the preset speech vector can be obtained by embedding the preset speech text segment, and there is an association relationship between the preset speech vector and the preset speech text segment. Since the target speech vector is determined from the preset speech vector in the speech vector library, the computer device can determine the target speech segment corresponding to the target speech vector from the preset speech text segment based on the association relationship between the preset speech vector and the preset speech text segment. Accordingly, the computer device can determine the target speech text segment as the search prompt information corresponding to the speech text to be retrieved, and when the computer device subsequently searches for the preset insurance case, it can perform a corresponding search based on the target speech text segment.

[0076] like Figure 2 As shown, since the preset speech text segment is obtained by semantic segmentation of the preset speech text, the text length of the preset speech text segment is less than the text segment of the preset speech text. In the case where the computer device uses the target speech text segment as the search prompt information, the probability of the target speech text segment appearing in the preset insurance report case can be higher than the probability of the target speech text appearing in the preset insurance report case, so that the target speech text segment is determined as the search prompt information, which can improve the search hit rate of the preset insurance report case, thereby facilitating the subsequent improvement of the convenience of searching the insurance report case.

[0077] Step S105: According to the search prompt information, determine the target insurance case corresponding to the to-be-searched speech text from a plurality of preset insurance cases.

[0078] For example, the computer device may determine whether the text corresponding to the search prompt information appears in multiple preset insurance reporting cases according to the instructions of the search prompt information. If the text corresponding to the search prompt information appears in the preset insurance reporting case, the computer device may determine that the similarity between the insurance reporting speech corresponding to the preset insurance reporting case and the speech text to be searched is greater than or equal to the preset similarity threshold, thereby determining the preset insurance reporting case as the target insurance reporting case corresponding to the speech text to be searched. The preset similarity threshold may be pre-set or set by the user, which is not limited here.

[0079] Take the insurance business as an example. When determining the target insurance case corresponding to the to-be-searched speech text, the business personnel can refer to the target insurance case to determine whether the insurance speech of the insurance personnel is insurance fraud speech. For another example, when it is necessary to verify the insurance speech corresponding to the unprocessed preset insurance case, the target insurance case corresponding to the to-be-searched speech text corresponding to the user input text can be determined from multiple preset insurance cases based on the user input text without the judgment of the business personnel, thereby realizing the verification of the insurance speech corresponding to multiple preset insurance cases.

[0080] In some embodiments, the search prompt information is input into a preset language model, so that the preset language model determines the target insurance case corresponding to the to-be-searched speech text from a plurality of preset insurance cases based on the search prompt information.

[0081] like Figure 2As shown, when the search prompt information is input into the preset language model, the preset language model can search for multiple preset insurance cases based on the search prompt information, and determine whether the search prompt information exists in each preset insurance case, so as to determine the similarity between the insurance report speech corresponding to the preset insurance case and the speech text to be searched, thereby determining the preset insurance case with a similarity greater than or equal to the preset similarity threshold as the target insurance case. When determining the target insurance case, the preset language model can output reply information to the user input text based on the target insurance case, which is not limited here. The preset language model can include a large language model, which is not limited here.

[0082] In this way, by determining the target insurance case corresponding to the to-be-searched speech text from multiple preset insurance cases based on the search prompt information, the target insurance case can assist business personnel in determining whether the insurance reporter's insurance speech is an insurance fraud speech, or it can determine the target insurance case whose insurance speech is similar to the insurance speech corresponding to the to-be-searched speech text without the need for business personnel to check multiple preset insurance cases, which is conducive to improving the convenience of searching for insurance cases.

[0083] The above-mentioned embodiment provides a method for retrieving insurance cases by obtaining a user input text; obtaining a to-be-retrieved speech text corresponding to the user input text; determining, based on the to-be-retrieved speech text, a target speech vector corresponding to the to-be-retrieved speech text from preset speech vectors included in a speech vector library; determining, based on the target speech vector, retrieval prompt information corresponding to the to-be-retrieved speech text; and, based on the retrieval prompt information, determining a target insurance case corresponding to the to-be-retrieved speech text from a plurality of preset insurance cases, so as to improve the convenience of retrieving insurance cases.

[0084] For example, it can be used for insurance business in the financial field. When handling insurance business, if the business personnel receive a call from the insurance personnel, they can determine the user input text according to the language description of the insurance personnel, and obtain the target insurance case related to the text to be retrieved corresponding to the user input text according to the retrieval method of the insurance case. The business personnel can refer to the target insurance case to determine whether the insurance personnel have committed insurance fraud, such as whether the language description of the insurance personnel involves insurance fraud speech. Accordingly, if the business personnel need to retrieve different target insurance cases corresponding to the same insurance speech, the business personnel can enter the relevant user input text, and obtain the target insurance case related to the text to be retrieved corresponding to the user input text through the retrieval method of the insurance case. Based on this, in the process of retrieving the target insurance case, there is no need to spend a lot of manpower and time to manually check multiple preset insurance cases, which is conducive to improving the convenience of retrieval of insurance cases, thereby helping to improve the processing efficiency of insurance claims business.

[0085] See also Figure 4 , Figure 4 This is a schematic block diagram of a retrieval device for insurance cases provided in an embodiment of the present application. The retrieval device can be configured in a server or an electronic device to execute the aforementioned retrieval method for insurance cases.

[0086] like Figure 4 As shown, the search device includes: a first text acquisition module 110, a second text acquisition module 120, a vector determination module 130, a search prompt determination module 140 and a case determination module 150.

[0087] The first text acquisition module 110 is used to acquire the text input by the user.

[0088] The second text acquisition module 120 is used to acquire the to-be-searched speech text corresponding to the user input text.

[0089] The vector determination module 130 is used to determine, based on the speech text to be retrieved, a target speech vector corresponding to the speech text to be retrieved from the preset speech vectors included in the speech vector library.

[0090] The search prompt determination module 140 is used to determine the search prompt information corresponding to the to-be-searched speech text according to the target speech vector.

[0091] The case determination module 150 is used to determine the target insurance case corresponding to the to-be-searched speech text from a plurality of preset insurance cases according to the search prompt information.

[0092] It should be noted that technicians in the relevant field can clearly understand that for the convenience and simplicity of description, the specific working processes of the above-described devices and modules and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0093] The method of the present application can be used in many general or special computer system environments or configurations. For example: personal computers, server computers, handheld devices or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronic devices, 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.

[0094] Exemplarily, the above method and apparatus may be implemented in the form of a computer program, which may be run on a computer device.

[0095] See also Figure 5 , Figure 5 1 is a schematic block diagram of a computer device provided in an embodiment of the present application. The computer device may be a server or an electronic device.

[0096] like Figure 5 As shown, the computer device includes a processor, a memory, and a network interface connected via a system bus, wherein the memory may include a storage medium and an internal memory.

[0097] The storage medium can store an operating system and a computer program. The computer program includes program instructions, and when the program instructions are executed, the processor can execute any method for retrieving insurance cases.

[0098] The processor is used to provide computing and control capabilities and support the operation of the entire computer equipment.

[0099] The internal memory provides an environment for the operation of the computer program in the storage medium. When the computer program is executed by the processor, the processor can execute any method for retrieving insurance cases.

[0100] The network interface is used for network communication, such as sending assigned tasks, etc. Those skilled in the art will appreciate that Figure 5 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0101] It should be understood that the processor may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0102] In one embodiment, the processor is used to execute a computer program and implement the following steps when executing the computer program:

[0103] Get user input text;

[0104] Obtain the to-be-searched speech text corresponding to the user input text;

[0105] According to the to-be-searched speech text, determine the target speech vector corresponding to the to-be-searched speech text from the preset speech vectors included in the speech vector library;

[0106] Determine, according to the target speech vector, the search prompt information corresponding to the speech text to be searched;

[0107] According to the search prompt information, a target insurance case corresponding to the to-be-searched speech text is determined from a plurality of preset insurance cases.

[0108] It should be noted that technical personnel in the relevant field can clearly understand that, for the convenience and brevity of description, the specific working process of the above-described retrieval of insurance cases can refer to the corresponding process in the aforementioned insurance case retrieval method embodiment, and will not be repeated here.

[0109] The embodiments of the present application also provide a computer-readable storage medium on which a computer program is stored. The method implemented when the computer program is executed by a processor can refer to the various embodiments of the method for retrieving insurance reports of the present application.

[0110] The computer-readable storage medium may be an internal storage unit of the computer device described in the foregoing embodiment, such as a hard disk or memory of the computer device. The computer-readable storage medium may also be an external storage device of the computer device, such as a plug-in hard disk, a smart memory card (SmartMedia Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc., equipped on the computer device.

[0111] It should be understood that the terms used in this application specification are only for the purpose of describing specific embodiments and are not intended to limit the application. As used in this application specification and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0112] It should also be understood that the term "and / or" used in the specification of this application and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, including these combinations. It should be noted that, in this article, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or system including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or system. In the absence of further restrictions, an element defined by the sentence "including a..." does not exclude the presence of other identical elements in the process, method, article or system including the element.

[0113] The serial numbers of the embodiments of the present application are for description only and do not represent the advantages and disadvantages of the embodiments. The above description is only a specific implementation mode of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed in the present application, and these modifications or replacements should be included in the protection scope of the present application. Therefore, the protection scope of the present application shall be based on the protection scope of the claims.

Claims

1. A method for retrieving insurance cases, characterized in that: include: Get user input text; Obtain the to-be-searched speech text corresponding to the user input text; According to the to-be-searched speech text, determine the target speech vector corresponding to the to-be-searched speech text from the preset speech vectors included in the speech vector library; Determine, according to the target speech vector, the search prompt information corresponding to the speech text to be searched; According to the search prompt information, a target insurance case corresponding to the to-be-searched speech text is determined from a plurality of preset insurance cases.

2. The search method according to claim 1, characterized in that: The step of determining, according to the to-be-retrieved speech text, a target speech vector corresponding to the to-be-retrieved speech text from preset speech vectors included in a speech vector library comprises: Embedding the to-be-retrieved speech text to obtain a to-be-retrieved speech vector corresponding to the to-be-retrieved speech text; Determine the similarity between the to-be-retrieved speech vector and a preset speech vector in the speech vector library; A preset speech vector whose similarity is greater than or equal to a preset similarity threshold is determined as a target speech vector corresponding to the speech text to be retrieved.

3. The search method according to claim 1, characterized in that: The step of determining, according to the target speech vector, search prompt information corresponding to the speech text to be searched includes: Based on the association relationship between the preset speech vector and the preset speech text segment, according to the target speech vector, determine the target speech text segment corresponding to the target speech vector from the preset speech text segment; The target speech text segment is determined as the search prompt information corresponding to the speech text to be searched.

4. The search method according to claim 1, characterized in that: Determining a target insurance case corresponding to the to-be-searched speech text from a plurality of preset insurance cases according to the search prompt information includes: The search prompt information is input into the preset language model, so that the preset language model determines the target insurance case corresponding to the to-be-searched speech text from a plurality of preset insurance cases according to the search prompt information.

5. The search method according to any one of claims 1 to 4, characterized in that: Before determining the target routine vector corresponding to the to-be-searched speech text from the preset speech vectors included in the speech vector library according to the to-be-searched speech text, the method further includes: Obtain the preset insurance reporting voice corresponding to the preset insurance reporting case; Obtaining the preset speech text corresponding to the preset insurance reporting voice; Performing semantic segmentation processing on the preset speech text to obtain preset speech text segments corresponding to the preset speech text; Embedding the preset speech text segment to obtain a preset speech vector corresponding to the preset speech text; A speech vector library is determined according to preset speech vectors corresponding to each of a plurality of preset speech texts.

6. The search method according to claim 5, characterized in that: The step of obtaining a preset speech text corresponding to the preset insurance reporting voice includes: Performing voice conversion processing on the preset insurance reporting voice to obtain a preset insurance reporting text corresponding to the preset insurance reporting voice; Performing data cleaning on the preset insurance report text to obtain an initial script text; According to the initial speech text and the speech label corresponding to the initial speech text, the preset speech text corresponding to the preset insurance reporting voice is determined.

7. The search method according to claim 5, characterized in that: The step of determining a speech vector library according to preset speech vectors corresponding to each of the plurality of preset insurance report texts includes: According to a preset data storage algorithm, a plurality of the preset speech vectors are stored to obtain a speech vector library.

8. A device for searching insurance cases, characterized in that: The retrieval device comprises: A first text acquisition module, used to acquire user input text; A second text acquisition module is used to acquire the to-be-searched speech text corresponding to the user input text; A vector determination module, used to determine, according to the to-be-retrieved speech text, a target speech vector corresponding to the to-be-retrieved speech text from preset speech vectors included in a speech vector library; A search prompt determination module, used to determine the search prompt information corresponding to the to-be-searched speech text according to the target speech vector; The case determination module is used to determine the target insurance case corresponding to the to-be-searched speech text from a plurality of preset insurance cases according to the search prompt information.

9. A computer device, characterized in that: The computer device includes a memory and a processor; The memory is used to store computer programs; The processor is used to execute the computer program and implement the insurance case retrieval method as described in any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the insurance case search method according to any one of claims 1 to 7 are implemented.

Images