Intelligent Insurance Product Speech Synthesis Method, Device, System and Storage Medium

Abstract

The present invention discloses an intelligent insurance product speech synthesis method, device, system, storage medium and computer device. The method includes: receiving a recording request sent by a client, where the recording request carries identification information of a recording task; querying, according to the identification information of the recording task, a first product corresponding to the recording task and n second products corresponding to the recording task, where n is greater than or equal to zero; obtaining the product speech of the first product and the product speech of the n second products; retaining the product speech of the first product, merging the links in the product speech of the n second products that are repeated with the product speech of the first product into the product speech of the first product, and appending the unmerged links to the merged speech to obtain the synthesized speech of the recording task. The above method ensures that each link in the merged product speech does not repeat each other, reduces the redundant content in the synthesized speech, reduces the recording duration, and improves the recording efficiency.

Description

Technical Field

[0001] The present invention relates to the technical field of computer information processing, and in particular, to an intelligent insurance product speech synthesis method, device, system, storage medium, and computer device. Background Art

[0002] According to the relevant management measures of the China Banking and Insurance Regulatory Commission, in the sales process of some insurance businesses, it is usually necessary to conduct dual recording of the entire sales process, that is, dual recording of audio and video of the sales process. Dual recording means that the business party collects audio-visual materials and electronic data by recording audio and video, and uses this to record and save the key links of the business sales process, so as to achieve important functions such as replayability of business sales behavior, queryability of important information, and confirmation of problem responsibility, in order to avoid non-compliant phenomena in the sales process.

[0003] In the current social environment, insurance product speech has gradually been upgraded from the manual recording method to the computer synthesis method. For example, in an insurance product speech synthesis method, the sales process of an insurance policy is segmented into several speech templates composed of speech segments according to the insurance product, and then the relevant information of the policyholder is added to each speech segment, and then each speech segment of each speech template is output in a preset order, and finally the speech segments of each speech template are converted into audio for output to be provided for users to use. This method effectively avoids many problems such as uncontrollable reading fluency of marketing personnel, uneven Putonghua levels, and non-standard business processes.

[0004] However, in the insurance industry, the insurance products provided by each insurance company are numerous and complex. This method of segmenting an insurance policy into several speech templates according to the product and then outputting each speech template in a set order will cause a large amount of redundancy in the speech content. Especially for the case where there are multiple insurance products in an insurance policy, it will greatly extend the recording duration, resulting in low recording efficiency and poor consumer experience. Summary of the Invention

[0005] In view of this, the present application provides an intelligent insurance product speech synthesis method, device, system, storage medium, and computer device, mainly aiming to solve the technical problems of redundant speech content, long recording time, and low recording efficiency in insurance product synthesis.

[0006] According to the first aspect of the present invention, an intelligent insurance product speech synthesis method is provided, and the method includes:

[0007] Receiving a recording request sent by a client, where the recording request carries identification information of a recording task;

[0008] Query the first product corresponding to the recording task and n second products corresponding to the recording task according to the identification information of the recording task, where n is greater than or equal to zero;

[0009] Obtain the product speech of the first product and the product speech of the n second products. The product speech consists of at least one link, and each link is configured with a link speech;

[0010] Retain the product speech of the first product, merge the links in the product speech of the n second products that are repeated with the product speech of the first product into the product speech of the first product, and append the unmerged links to the merged speech to obtain the synthesized speech of the recording task.

[0011] Optionally, the recording request also carries the task parameters and product parameters of the recording task; then after obtaining the product speech of the first product and the product speech of the n second products, the method further includes: mapping the task parameters and product parameters of the recording task to the link speeches of each link of the product speech of the first product and the product speech of the n second products respectively; verifying each parameter in the link speech of each link through a regular expression, and saving the verified product speech of the first product and the product speech of the n second products.

[0012] Optionally, merging the links in the product speech of the n second products that are repeated with the product speech of the first product into the product speech of the first product and appending the unmerged links to the merged speech includes: comparing each link of the product speech of the n second products with each link of the product speech of the first product in turn, and finding the links in the product speech of the n second products that are repeated with the product speech of the first product; merging the links in the product speech of the n second products that have the merge attribute and are repeated with the product speech of the first product into the corresponding link of the product speech of the first product; appending the links in the product speech of the n second products that do not have the merge attribute and the links that are not repeated with the product speech of the first product to the merged speech.

[0013] Optionally, after appending the unmerged links to the merged speech, the method further includes: obtaining the head speech and tail speech of each link, and adding the head speech to the head position of the link speech of the corresponding link, and adding the tail speech to the tail position of the link speech of the corresponding link.

[0014] Optionally, the method further includes: generating speech text content according to the synthesized speech of the recording task, and storing the synthesized speech of the recording task and the speech text content in a database.

[0015] Optionally, the method further includes: performing speech synthesis on the synthesized speech of the recording task to obtain the audio information of the recording task.

[0016] Optionally, perform speech synthesis on the synthesized speech of the recording task to obtain the audio information of the recording task, including: receiving an audio synthesis request sent by the client, where the audio synthesis request carries a speech rate parameter and a voice color parameter; obtaining the speech text content of the recording task, and converting the speech text content into audio information according to the speech rate parameter and the voice color parameter.

[0017] Optionally, before receiving the recording request sent by the client, this method further includes: configuring each link and the link speech of each link, and configuring the product speech of each product according to each link and the link speech of each link.

[0018] According to the second aspect of the present invention, there is provided an intelligent insurance product speech synthesis device, which includes:

[0019] A request acquisition module, configured to receive a recording request sent by the client, where the recording request carries identification information of the recording task;

[0020] A product query module, configured to query the first product corresponding to the recording task and the n second products corresponding to the recording task according to the identification information of the recording task, where n is greater than or equal to zero;

[0021] A speech acquisition module, configured to acquire the product speech of the first product and the product speech of the n second products, where the product speech is composed of at least one link, and each link is configured with a link speech;

[0022] A speech synthesis module, configured to retain the product speech of the first product, merge the links in the product speech of the n second products that are repeated with the product speech of the first product into the product speech of the first product, and append the unmerged links to the merged speech to obtain the synthesized speech of the recording task.

[0023] According to the third aspect of the present invention, there is provided an intelligent insurance product speech synthesis system, which includes:

[0024] A client, configured to send a recording request, where the recording request carries identification information of the recording task;

[0025] A server, configured to receive the recording request, and query the first product corresponding to the recording task and the n second products corresponding to the recording task according to the identification information of the recording task in the recording request, where n is greater than or equal to zero;

[0026] Acquire the product speech of the first product and the product speech of the n second products, where the product speech is composed of at least one link, and each link is configured with a link speech;

[0027] Retain the product script of the first product, merge the parts in the product scripts of the n second products that are repeated with the product script of the first product into the product script of the first product, and append the unmerged parts to the merged script to obtain the synthesized script for the recording task.

[0028] According to the fourth aspect of the present invention, there is provided a storage medium on which a computer program is stored, and when the program is executed by a processor, the above-mentioned intelligent insurance product script synthesis method is implemented.

[0029] According to the fifth aspect of the present invention, there is provided a computer device including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, the above-mentioned intelligent insurance product script synthesis method is implemented.

[0030] An intelligent insurance product script synthesis method, device, storage medium and computer device provided by the present invention first receive a recording request sent by a client, then query the first product and n second products corresponding to the recording task according to the identification information of the recording task in the recording request, and then obtain the product scripts of the first product and the n second products. Finally, retain the product script of the first product according to the link structure of the first product, merge the parts in the product scripts of the n second products that are repeated with the product script of the first product into the product script of the first product, and append the unmerged parts to the merged script to obtain the synthesized script for the recording task. By merging the repeated parts in the product scripts into the same part, the above method ensures that the parts in the merged product script do not repeat each other, reduces the redundant content in the synthesized product script, reduces the recording duration, and improves the recording efficiency. Moreover, the above method also retains the product script of the first product and then appends the differentiated parts of the n second products, so that the recorded content is not omitted, and the synthesized product script is well-organized and easy to be accepted by consumers, improving the consumer experience.

[0031] The above description is only an overview of the technical solution of this application. In order to be able to understand the technical means of this application more clearly, it can be implemented according to the content of the specification. And in order to make the above and other purposes, features and advantages of this application more obvious and understandable, the specific embodiments of this application are specifically given below. Description of the Drawings

[0032] The drawings described herein are used to provide a further understanding of the present invention, and constitute a part of this application. The schematic embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation to the present invention. In the drawings:

[0033] Figure 1Shows a schematic flowchart of a method for synthesizing intelligent insurance product sales scripts provided by an embodiment of the present invention;

[0034] Figure 2 Shows a schematic flowchart of another method for synthesizing intelligent insurance product sales scripts provided by an embodiment of the present invention;

[0035] Figure 3 Shows a schematic structural diagram of an intelligent insurance product sales script synthesis device provided by an embodiment of the present invention;

[0036] Figure 4 Shows a schematic structural diagram of another intelligent insurance product sales script synthesis device provided by an embodiment of the present invention;

[0037] Figure 5 Shows a schematic structural diagram of an intelligent insurance product sales script synthesis system provided by an embodiment of the present invention. Detailed implementation manners

[0038] The present invention will be described in detail below with reference to the accompanying drawings and in conjunction with embodiments. It should be noted that, without conflict, the embodiments in the present application and the features in the embodiments can be combined with each other.

[0039] In one embodiment, as Figure 1 shown, a method for synthesizing intelligent insurance product sales scripts is provided. Taking the application of this method to computer devices such as a server as an example, the method includes the following steps:

[0040] 101. Receive a recording request sent by the client, where the recording request carries identification information of the recording task.

[0041] The client herein refers to a device equipped with a camera device and a recording device. For example, the client can be a smart tablet, a computer, a smart mobile phone, a smart interaction robot, etc. Specifically, the user can log in to relevant software through the client and then send a recording request to the server through this software. In this embodiment, each recording request corresponds to a recording task, and each recording task corresponds to a unique identification information. Through this identification information, the server can query all the product information corresponding to this recording task. It can be understood that the recording task is customized and initiated by the user. Therefore, each recording task can correspond to different product combinations, where the products referred to in this embodiment can specifically be insurance products.

[0042] 102. According to the identification information of the recording task, query the first product corresponding to the recording task and n second products corresponding to the recording task, where n is greater than or equal to zero.

[0043] Among them, the first product is also known as the main product or main policy, which refers to the product that must be included in a recording task. The second product is also known as the sub-product or sub-policy, which refers to the other products except the first product in a recording task. The second product can be one, multiple, or none. The first product and the second product do not affect each other. In addition, the first product can be determined by the system or specified by the user. For example, the server can determine that the first added product is the first product, or set the product specified by the user as the first product.

[0044] Specifically, the server can query the first product and n second products corresponding to the recording task in the database according to the identification information of the recording task, where n is an integer greater than or equal to zero. That is, the server can obtain all the products corresponding to the recording task according to the identification information of the recording task, and all the products include one first product and zero or more than zero second products.

[0045] 103. Obtain the product speech of the first product and the product speeches of n second products. The product speech consists of at least one link, and each link is configured with a link speech.

[0046] Among them, each product has a complete set of product speeches, and the product speech of each product consists of at least one link, and each link is configured with a link speech. Specifically, each link has a unique identifier, a link name, a merge attribute, a head speech content, and a tail speech content. Further, the link speech refers to the speech content set for the link. The link speech corresponding to each link can be one or more, but only one of them can be selected when configuring the link speech for the link of the product speech. The link speech can include various contents such as link content, link speech parameters, link speech judgment, and link speech actions.

[0047] Specifically, the server can obtain the product speeches of each product corresponding to the recording task, including the product speech of the first product and the product speeches of n second products. In this way, the server can obtain all the links corresponding to the product speech of the first product and all the links corresponding to the product speeches of n second products. Among them, the links corresponding to the product speech of each product do not repeat each other, but the links between different products may be repeated. For example, the first product and a second product both have two identical links, and the link names are "Seek the opinion of the applicant" and "Remind the applicant to pay attention" respectively. Then for these two links, they are the links in the product speech of the second product that are repeated with the product speech of the first product.

[0048] 104. Retain the product pitch of the first product, merge the parts in the product pitches of the n second products that are repeated with the product pitch of the first product into the product pitch of the first product, and append the unmerged parts to the merged pitch to obtain the synthesized pitch for the recording task.

[0049] Specifically, the product pitch of the first product obtained can be combined into a whole according to its link structure. Here, the link structure refers to all the links included in the product pitch of a product and the sequence of each link. That is, the server can first retain the product pitch of the first product, and then merge the parts in the product pitches of the n second products that are repeated with the product pitch of the first product into the product pitch of the first product. When merging, first determine whether the link of the second product is repeated with the link of the first product, and then determine whether the repeated link has a merge attribute. If a link of the second product is both repeated with the link of the first product and has a merge attribute, then merge this link into the product pitch of the first product. Finally, append all the remaining different links to the product pitch of the first product in sequence. For example, all the different links of the n second products can be appended to the back of the pitch content of the first product in sequence. In this way, it can not only ensure that the pitch content of the first product is clear and coherent, but also ensure that the corresponding link pitches and product pitches of the second products merged later will not appear repeatedly, reducing the number of links and the recording duration of the product synthesized pitch.

[0050] The intelligent insurance product pitch synthesis method provided in this embodiment first receives the recording request sent by the client, then queries the first product and the n second products corresponding to the recording task according to the identification information of the recording task in the recording request, then obtains the product pitch of the first product and the product pitches of the n second products, and finally retains the product pitch of the first product according to the link structure of the first product, merges the parts in the product pitches of the n second products that are repeated with the product pitch of the first product into the product pitch of the first product, and append the unmerged parts to the merged pitch to obtain the synthesized pitch for the recording task. By merging the repeated links in the product pitch into the same link, the above method ensures that the links in the merged product pitch do not repeat each other, reduces the redundant content in the product synthesized pitch, reduces the recording duration, and improves the recording efficiency. Moreover, by retaining the product pitch of the first product and then appending the different links of the n second products, the above method ensures that the recording content is not omitted, makes the synthesized product pitch well-organized and easy to be accepted by consumers, and improves the consumer experience.

[0051] Further, as a refinement and extension of the specific implementation manner of the above embodiment, to fully illustrate the implementation process of this embodiment, an intelligent insurance product pitch synthesis method is provided, as Figure 2As shown, the method includes the following steps:

[0052] 201. Configure each link and the link script for each link, and configure the product script for each product according to each link and the link script for each link.

[0053] Specifically, before synthesizing the script, the server can pre-configure each link required for the product script, the link script for each link, and the product script for each product. During the process of configuring the link, the unique identifier, link name, and merge attribute of each link can be created first, and the head script (such as the guiding language) and tail script (such as the ending language) of each link can be created. In addition, in this embodiment, all links can be stored in the same space (for example, this space can be called the standard script link pool), and it is ensured that all link scripts in this space do not repeat each other, thereby improving the acquisition efficiency of the link script.

[0054] Furthermore, after configuring the unique identifier, link name, and merge attribute of each link, the link script for each link can be further configured. The link script corresponding to each link can be one or more. Among them, the link script can include link content, link script parameters, link script judgment, link script actions, and link keywords, etc. Among them, the link script actions can be set to photograph certificates, photograph documents, standard responses, key actions, and waiting, etc. After configuring the link script, the product script can be further configured. Configuring the product script includes configuring the applicable scope of the product script and configuring each link corresponding to the product script and the link script corresponding to each link. Among them, configuring the applicable scope can include configuring sales channels, task types, product parameters, product attributes, and product names, etc. For each link corresponding to the product script and the link script corresponding to each link, the links and the sequence of each link can be configured for the product script first, and then the link script of each link can be configured into the corresponding link to obtain the product script. In addition, the server can also configure quality inspection standard parameters, such as configuring the video in-frame ratio, keyword, and action response situation, etc. This quality inspection standard parameter can be used to check whether the recorded video is compliant, so as to facilitate subsequent verification and assist manual quality inspection.

[0055] 202. Receive the recording request sent by the client, where the recording request carries the identification information of the recording task, the task parameters of the recording task, and the product parameters.

[0056] Among them, the client refers to a device equipped with a camera device and a recording device. For example, the client can be a smart tablet, a computer, a smart mobile phone, a smart interaction robot, etc. Specifically, the user can log in to the corresponding software through the client, and then send a recording request to the server through the software. In this embodiment, each recording request corresponds to a recording task, and each recording task corresponds to a unique identification information. Through this identification information, the server can query all product information corresponding to the recording task. It can be understood that the recording task is customized and initiated by the user. Therefore, each recording task can correspond to different product combinations. Among them, the products referred to in this embodiment can specifically be insurance products.

[0057] In addition, the recording request sent by the client can also carry the task information of the recording task, including task parameters and product parameters of the recording task and other information. Among them, the task parameters refer to parameter information related to the person of the product, such as the practicing certificate information and ID card information of the applicant, and the product parameters refer to parameter information related to the product, such as the premium payment method and premium of the insurance policy and other information. Specifically, after the client creates the recording task, the user can fill in the relevant task information and add the relevant product information in the corresponding software of the client, and then the client sends a recording request to the server and carries the identification information of the recording task, the task parameters and product parameters of the recording task in the recording request.

[0058] 203. Query the first product corresponding to the recording task and n second products corresponding to the recording task according to the identification information of the recording task, where n is greater than or equal to zero.

[0059] Among them, the first product is also called the main product or the main insurance policy, which refers to the product that must be included in a recording task. The second task is also called the sub-product or the sub-insurance policy, which refers to the other products except the first product in a recording task. The second product can be one, or multiple, or none. The first product and the second product do not affect each other. In addition, the first product can be determined by the system or specified by the user. For example, the server can determine that the first added product is the first product, or can set the product specified by the user as the first product.

[0060] Specifically, the server can query the first product corresponding to the recording task and n second products in the database according to the identification information of the recording task, where n is an integer greater than or equal to zero. That is, the server can obtain all products corresponding to the recording task according to the identification information of the recording task. The all products include a first product and zero or more than zero second products.

[0061] 204. Obtain the product scripts of the first product and the product scripts of n second products. The product script consists of at least one segment, and each segment is configured with a segment script.

[0062] Among them, each product has a complete set of product scripts. The product script of each product consists of at least one segment, and each segment is configured with a segment script. Specifically, each segment has a unique identifier, a segment name, a merge attribute, a head script content, and a tail script content. Further, the segment script refers to the script content set for the segment. The segment script corresponding to each segment can be one or more, but only one of them can be selected when configuring the segment script for the segment of the product script. Among them, the segment script can contain various contents such as segment content, segment script parameters, segment script judgment, and segment script actions.

[0063] Specifically, the server can obtain the product scripts of each product corresponding to the recording task, including the product script of the first product and the product scripts of n second products. In this way, the server can obtain all the segments corresponding to the product script of the first product and all the segments corresponding to the product scripts of n second products. Among them, the segments corresponding to the product script of each product do not repeat each other, but the segments between different products may be repeated. For example, the first product and a second product both have two identical segments, and the segment names are "Seek the opinion of the applicant" and "Remind the applicant to pay attention" respectively. Then for these two segments, they are the segments in the product script of the second product that are repeated with the product script of the first product.

[0064] 205. Map the task parameters and product parameters of the recording task to the segment scripts of each segment of the product script of the first product and the product scripts of n second products respectively.

[0065] Specifically, after obtaining the product script of the first product and the product scripts of n second products, the server can map the task parameters and product parameters carried in the recording request to the segment scripts of each segment of the product scripts of each product one by one. When performing parameter mapping, the task parameters and product parameters can be mapped to the segment scripts of each segment corresponding to the product script of each product in sequence according to the order of the first product and n second products, and the task parameters and product parameters are used to replace the segment script parameters in the segment script until all products are mapped.

[0066] 206. Verify each parameter in the segment script of each segment through regular expressions, and save the verified product script of the first product and the product scripts of n second products.

[0067] Specifically, after mapping the task parameters and product parameters, the server can also parse and verify the task parameters and product parameters in the conversation scripts of each link through regular expressions, so as to verify whether the formats and contents of the parameters mapped to the conversation scripts of each link are correct. If all parameters pass the verification, it can be considered that the filling of the conversation script content is completed. Then, the server can save each filled product conversation script independently for subsequent conversation script merging.

[0068] 207. Retain the product conversation script of the first product, merge the links in the product conversation scripts of the n second products that are repeated with the product conversation script of the first product into the product conversation script of the first product, and append the unmerged links to the merged conversation script.

[0069] Specifically, the obtained product conversation script of the first product can be combined into a whole according to its link structure. Here, the link structure refers to all the links included in the product conversation script of a product and the sequence of each link. That is, the server can first retain the product conversation script of the first product, and then merge the links in the product conversation scripts of the n second products that are repeated with the product conversation script of the first product into the product conversation script of the first product. When merging, it is necessary to first determine whether the link of the second product is repeated with the link of the first product, and then determine whether the repeated link has a merge attribute set. If a link of the second product is both repeated with the link of the first product and has a merge attribute set, then merge this link into the product conversation script of the first product. Finally, append all the remaining different links to the product conversation script of the first product in sequence. For example, all the different links of the n second products can be appended to the rear side of the conversation script content of the first product in sequence. In this way, it can not only ensure the clarity and coherence of the conversation script content of the first product, but also ensure that the corresponding link conversation scripts and product conversation scripts of the second products to be merged later will not appear repeatedly, reducing the number of links and recording duration of the product composite conversation script.

[0070] In an optional implementation manner of this embodiment, the specific method of merging the links in the product conversation scripts of the n second products that are repeated with the product conversation script of the first product into the product conversation script of the first product and appending the unmerged links to the merged conversation script includes the following steps: First, compare each link of the product conversation scripts of the n second products with each link of the product conversation script of the first product in sequence, and find the links in the product conversation scripts of the n second products that are repeated with the product conversation script of the first product. Then, merge the links in the product conversation scripts of the n second products that have a merge attribute set and are repeated with the product conversation script of the first product into the corresponding links of the product conversation script of the first product. Finally, append the links in the product conversation scripts of the n second products that do not have a merge attribute set and are not repeated with the product conversation script of the first product to the merged conversation script.

[0071] 208. Obtain the head and tail scripts for each link, add the head script to the head position of the link script for the corresponding link, and add the tail script to the tail position of the link script for the corresponding link to obtain the synthesized script for the recording task.

[0072] Specifically, when the server obtains the product scripts of the first product and the product scripts of n second products, it can also obtain the head and tail scripts for each link. Then, after merging the scripts of the first product and the n second products, add the head scripts of each link to the head position of the link script for the corresponding link, and add the tail scripts to the tail position of the link script for the corresponding link, so as to obtain the synthesized script for the final recording task. For example, for the link "Seek the opinion of the policyholder", the head script for this link is "Hello", and the tail script is "Okay, thank you". Then, in this step, "Hello" can be added to the head position of the link script for this link, and "Okay, thank you" can be added to the tail position of the link script for this link. It should be noted that the head and tail scripts corresponding to each link are configured according to the characteristics of this link. For example, for the link "Remind the policyholder to pay attention", its head script is "Please note", and the tail script is "The playback of the attention content is completed".

[0073] 209. Generate the script text content according to the synthesized script of the recording task, and store the synthesized script and the script text content of the recording task in the database.

[0074] Specifically, the server can generate the script text content of the recording task according to the synthesized script of the recording task, and then can also store the synthesized script and the script text content of the recording task in the database. When storing, the server can specify the storage locations of the synthesized script and the script text content for subsequent quality inspection. In addition, the generated script text content can also be cached in the server first for subsequent audio synthesis.

[0075] 210. Perform voice synthesis on the synthesized script of the recording task to obtain the audio information of the recording task.

[0076] Specifically, the server can convert the synthesized speech of the recording task into audio information in combination with the speech text content and send the audio information to the client. After receiving the audio information of the recording task, the client can synchronously play the audio information of the synthesized speech when recording a video. Moreover, during the process of playing the audio information, the client can also determine whether a segment speech action is inserted according to the speech text content. If a segment speech action is inserted, corresponding action recognition and detection can be performed according to the action type and action parameters of the segment speech action. For example, identify document information through object recognition technology, determine whether text documents such as insurance policies and contracts are displayed through target detection technology, identify whether the content of the user's answer is the same as the set response content through keyword recognition technology and speech-to-text technology, and detect whether the user has completed key actions such as signing through target detection technology and action recognition technology, etc. In addition, the audio file can also respond to the volume button to control the playback volume, and respond to the play button to control the playback and pause of the audio, etc.

[0077] In an optional implementation manner, step 210 can be specifically implemented as follows: First, receive an audio synthesis request sent by the client, where the audio synthesis request carries speech rate parameters (such as steady, fluent, rapid) and voice color parameters (such as male voice, female voice). Then, the server converts the obtained speech text content into audio content according to the speech rate parameters and voice color parameters set by the user to obtain the audio information of the recording task.

[0078] The intelligent insurance product speech synthesis method provided in this embodiment enhances the dynamic configurability of speech synthesis by converting the traditional template configuration type into a segment configuration type, effectively reducing redundant segment speeches. Especially when the recording task includes multiple products, it not only retains the integrity of the speech but also improves the efficiency of speech recording and the user experience. In addition, the speech synthesis method provided in this embodiment has good flexibility and can further improve the user experience.

[0079] Furthermore, as Figure 1 、 Figure 2 [[ID=?]] Figure 3 shown, the device includes: a request acquisition module 31, a product query module 32, a speech acquisition module 33, and a speech synthesis module 34.

[0080] The request acquisition module 31 is configured to receive a recording request sent by the client, where the recording request carries identification information of the recording task;

[0081] It should be noted that there seems to be an error in the original text where the "?" in line 13 should be a correct reference number. Please check and correct it if possible.The product query module 32 can be used to query the first product corresponding to the recording task and the n second products corresponding to the recording task according to the identification information of the recording task, where n is greater than or equal to zero;

[0082] The script acquisition module 33 can be used to acquire the product scripts of the first product and the n second products, where the product script consists of at least one link, and each link is configured with a link script;

[0083] The script synthesis module 34 can be used to retain the product script of the first product, merge the links in the product scripts of the n second products that are repeated with the product script of the first product into the product script of the first product, and append the unmerged links to the merged script to obtain the synthesized script of the recording task.

[0084] In a specific application scenario, the recording request also carries the task parameters and product parameters of the recording task, then as Figure 4 shown, the device further includes a parameter mapping module 35 and a parameter verification module 36. The parameter mapping module 35 can specifically be used to map the task parameters and product parameters of the recording task to the link scripts of each link in the product script of the first product and the product scripts of the n second products respectively; the parameter verification module 36 can specifically be used to verify each parameter in the link scripts of each link through regular expressions, and save the verified product script of the first product and the product scripts of the n second products.

[0085] In a specific application scenario, the script synthesis module 34 can specifically be used to compare each link of the product scripts of the n second products with each link of the product script of the first product in turn, and find the links in the product scripts of the n second products that are repeated with the product script of the first product; merge the links in the product scripts of the n second products that have the merge attribute and are repeated with the product script of the first product into the corresponding links of the product script of the first product; append the links in the product scripts of the n second products that do not have the merge attribute and are not repeated with the product script of the first product to the merged script.

[0086] In a specific application scenario, the script synthesis module 34 can also be used to acquire the head script and tail script of each link, and add the head script to the head position of the link script of the corresponding link, and add the tail script to the tail position of the link script of the corresponding link.

[0087] In a specific application scenario, as Figure 4 shown, the device further includes a script storage module 37. The script storage module 37 can specifically be used to generate script text content according to the synthesized script of the recording task, and store the synthesized script of the recording task and the script text content in the database.

[0088] In a specific application scenario, such as Figure 4 shown, the device further includes a speech synthesis module 38, and the speech synthesis module 38 can specifically be used to perform speech synthesis on the synthesized words of the recording task to obtain the audio information of the recording task.

[0089] In a specific application scenario, the speech synthesis module 38 can specifically be used to receive an audio synthesis request sent by a client, where the audio synthesis request carries a speech rate parameter and a voice color parameter; obtain the text content of the words of the recording task, and convert the text content of the words into audio information according to the speech rate parameter and the voice color parameter.

[0090] In a specific application scenario, such as Figure 4 shown, the device further includes a words configuration module 39, and the words configuration module 39 can specifically be used to configure each link and the link words of each link, and configure the product words of each product according to each link and the link words of each link.

[0091] It should be noted that for other corresponding descriptions of each functional unit involved in the intelligent insurance product words synthesis device provided in this embodiment, reference can be made to Figure 1 、 Figure 2 the corresponding descriptions therein, which will not be elaborated here.

[0092] Furthermore, as the specific implementation of the method shown in Figure 1 、 Figure 2 and the device shown in Figure 3 、 Figure 4 this embodiment also provides an intelligent insurance product words synthesis system, as Figure 5 shown, the system includes: a client 41 and a server 42, where:

[0093] The client 41 can be used to send a recording request, where the recording request carries the identification information of the recording task;

[0094] The server 42 can be used to receive the recording request, and according to the identification information of the recording task in the recording request, query the first product corresponding to the recording task and n second products corresponding to the recording task, where n is greater than or equal to zero; obtain the product words of the first product and the product words of the n second products, where the product words are composed of at least one link, and each link is configured with a link word; retain the product words of the first product, merge the links in the product words of the n second products that are repeated with the product words of the first product into the product words of the first product, and append the unmerged links to the merged words to obtain the synthesized words of the recording task.

[0095] In a specific application scenario, the recording request also carries the task parameters and product parameters of the recording task. Then, the server 42 can also be used to map the task parameters and product parameters of the recording task to the segment scripts of the product script of the first product and the product scripts of n second products respectively; verify each parameter in the segment scripts of each segment through regular expressions, and save the verified product script of the first product and the product scripts of n second products.

[0096] In a specific application scenario, the server 42 can specifically be used to compare each segment of the product scripts of n second products with each segment of the product script of the first product in sequence, and find the segments in the product scripts of n second products that are repeated with the product script of the first product; merge the segments in the product scripts of n second products that have the merge attribute set and are repeated with the product script of the first product into the corresponding segments of the product script of the first product; append the segments in the product scripts of n second products that do not have the merge attribute set and the segments that are not repeated with the product script of the first product to the merged script.

[0097] In a specific application scenario, the server 42 can also be used to obtain the head script and tail script of each segment, and add the head script to the head position of the segment script of the corresponding segment, and add the tail script to the tail position of the segment script of the corresponding segment.

[0098] In a specific application scenario, the server 42 can also be used to generate script text content according to the synthesized script of the recording task, and store the synthesized script and the script text content of the recording task in the database.

[0099] In a specific application scenario, the server 42 can also be used to perform voice synthesis on the synthesized script of the recording task to obtain the audio information of the recording task.

[0100] In a specific application scenario, the server 42 can specifically also be used to receive an audio synthesis request sent by the client. Among them, the audio synthesis request carries a speech rate parameter and a voice color parameter; obtain the script text content of the recording task, and convert the script text content into audio information according to the speech rate parameter and the voice color parameter.

[0101] In a specific application scenario, the server 42 can also be used to configure each segment and the segment script of each segment, and configure the product script of each product according to each segment and the segment script of each segment.

[0102] It should be noted that for other corresponding descriptions of each functional unit involved in an intelligent insurance product script synthesis system provided in this embodiment, reference can be made to Figure 1 、 Figure 2 the corresponding descriptions therein, which will not be elaborated here.

[0103] Based on the above as Figure 1 、 Figure 2 shown in the method, correspondingly, this embodiment also provides a storage medium, on which a computer program is stored, and when the program is executed by a processor, it implements the intelligent insurance product speech synthesis method as shown in the above as Figure 1 、 Figure 2 shown.

[0104] Based on such an understanding, the technical solution of this application can be embodied in the form of a software product. This software product to be recognized can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.), and includes several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in various implementation scenarios of this application.

[0105] Based on the above as Figure 1 、 Figure 2 shown in the method, and Figure 3 and Figure 4 shown in the embodiment of the intelligent insurance product speech synthesis device, for the purpose of achieving the above object, this embodiment also provides an entity device for intelligent insurance product speech synthesis, which can specifically be a personal computer, a server, a smart phone, a tablet computer, a smart watch, or other network devices, etc. This entity device includes a storage medium and a processor; the storage medium is used to store a computer program; the processor is used to execute the computer program to implement the method as shown in the above as Figure 1 、 Figure 2 shown.

[0106] Optionally, this entity device may further include a user interface, a network interface, a camera, a Radio Frequency (RF) circuit, sensors, an audio circuit, a WI-FI module, etc. The user interface may include a display screen (Display), an input unit such as a keyboard (Keyboard), etc. Optionally, the user interface may further include a USB interface, a card reader interface, etc. The network interface may optionally include a standard wired interface, a wireless interface (such as a WI-FI interface), etc.

[0107] Those skilled in the art can understand that the structure of an entity device for intelligent insurance product speech synthesis provided in this embodiment does not constitute a limitation on this entity device, and it may include more or fewer components, or combine some components, or arrange different components.

[0108] The storage medium may further include an operating system and a network communication module. The operating system is a program for managing the hardware of the above-mentioned physical devices and the software resources to be recognized, and supports the operation of information processing programs and other software and / or programs to be recognized. The network communication module is used to implement communication between components inside the storage medium, as well as communication with other hardware and software in the information processing physical device.

[0109] Through the description of the above embodiments, those skilled in the art can clearly understand that the present application can be implemented by means of software plus a necessary general hardware platform, or can also be implemented by hardware. By applying the technical solution of the present application, first, a recording request sent by the client is received, and then according to the identification information of the recording task in the recording request, the first product and n second products corresponding to the recording task are queried. Subsequently, the product speech of the first product and the product speeches of the n second products are obtained. Finally, the product speech of the first product is retained according to the link structure of the first product, and the links in the product speeches of the n second products that are repeated with the product speech of the first product are merged into the product speech of the first product, and the unmerged links are appended to the merged speech to obtain the synthesized speech of the recording task. Compared with the prior art, the above method merges the repeated links in the product speech into the same link, ensuring that the links in the merged product speech do not repeat each other, reducing the redundant content in the synthesized product speech, reducing the recording duration, and improving the recording efficiency. Moreover, the above method also retains the product speech of the first product and then appends the differentiated links of the n second products, so that the recorded content is not omitted, and the synthesized product speech is well-organized and easy to be accepted by consumers, improving the consumer experience.

[0110] Those skilled in the art can understand that the drawings are only schematic diagrams of a preferred embodiment scenario, and the modules or processes in the drawings are not necessarily essential for implementing the present application. Those skilled in the art can understand that the modules in the device in the embodiment scenario can be distributed in the device in the embodiment scenario according to the description of the embodiment scenario, or can be correspondingly changed and located in one or more devices different from the present embodiment scenario. The modules in the above embodiment scenario can be combined into one module, or can be further split into multiple sub-modules.

[0111] The above serial numbers of the present application are only for description and do not represent the superiority or inferiority of the embodiment scenarios. The above disclosure only shows several specific embodiment scenarios of the present application. However, the present application is not limited thereto, and any changes that can be thought of by those skilled in the art should fall within the protection scope of the present application.

Claims

1. An intelligent insurance product speech synthesis method, characterized in that, The method includes: Receiving a recording request sent by a client, where the recording request carries identification information of a recording task; Querying, according to the identification information of the recording task, a first product corresponding to the recording task and n second products corresponding to the recording task, where n is greater than zero; Obtaining the product script of the first product and the product scripts of the n second products, where the product script consists of at least one link, and each link is configured with a link script; Retaining the product script of the first product, merging the links in the product scripts of the n second products that are repeated with the product script of the first product into the product script of the first product, and appending the unmerged links to the merged script to obtain the synthesized script of the recording task. Specifically, it includes: comparing each link of the product scripts of the n second products with each link of the product script of the first product in sequence, and finding the links in the product scripts of the n second products that are repeated with the product script of the first product; merging the links in the product scripts of the n second products that have the merge attribute and are repeated with the product script of the first product into the corresponding links of the product script of the first product; appending the links in the product scripts of the n second products that do not have the merge attribute and the links that are not repeated with the product script of the first product to the merged script.

2. The method according to claim 1, wherein The recording request also carries task parameters and product parameters of the recording task; then after obtaining the product script of the first product and the product scripts of the n second products, the method further includes: Mapping the task parameters and product parameters of the recording task to the link scripts of each link of the product script of the first product and the product scripts of the n second products respectively; Validating each parameter in the link scripts of each link through a regular expression, and saving the validated product script of the first product and the product scripts of the n second products.

3. The method according to claim 1, wherein After appending the unmerged links to the merged script, the method further includes: Obtaining the head script and tail script of each link, adding the head script to the head position of the link script of the corresponding link, and adding the tail script to the tail position of the link script of the corresponding link.

4. The method according to claim 1, characterized in that The method further includes: Generating script text content according to the synthesized script of the recording task, and storing the synthesized script and the script text content of the recording task in a database.

5. The method according to claim 3, characterized in that, The method further includes: Performing speech synthesis on the synthesized script of the recording task to obtain audio information of the recording task.

6. The method according to claim 5, wherein The performing speech synthesis on the synthesized script of the recording task to obtain audio information of the recording task includes: Receiving an audio synthesis request sent by a client, where the audio synthesis request carries a speech rate parameter and a voice color parameter; Obtaining the script text content of the recording task, and converting the script text content into audio information according to the speech rate parameter and the voice color parameter.

7. The method according to claim 1, characterized in that Before receiving the recording request sent by the client, the method further includes: Configure each link and the link script for each link, and configure the product script for each product according to the above-mentioned each link and the link script for each link.

8. An intelligent insurance product speech synthesis device, characterized in that The device includes: A request acquisition module, configured to receive a recording request sent by a client, where the recording request carries identification information of a recording task; A product query module, configured to query a first product corresponding to the recording task and n second products corresponding to the recording task according to the identification information of the recording task, where n>0; A script acquisition module, configured to acquire the product script of the first product and the product scripts of the n second products, where the product script consists of at least one link, and each link is configured with a link script; A script synthesis module, configured to retain the product script of the first product, merge the links in the product scripts of the n second products that are repeated with the product script of the first product into the product script of the first product, and append the unmerged links to the merged script to obtain the synthesized script of the recording task; Specifically, the script synthesis module is configured to compare each link of the product scripts of the n second products with each link of the product script of the first product in sequence, and find the links in the product scripts of the n second products that are repeated with the product script of the first product; merge the links in the product scripts of the n second products that have the merge attribute and are repeated with the product script of the first product into the corresponding links of the product script of the first product; append the links in the product scripts of the n second products that do not have the merge attribute and are not repeated with the product script of the first product to the merged script.

9. An intelligent insurance product speech synthesis system, characterized in that, The system includes: A client, configured to send a recording request, where the recording request carries identification information of a recording task; A server, configured to receive the recording request, and query a first product corresponding to the recording task and n second products corresponding to the recording task according to the identification information of the recording task in the recording request, where n>0; Acquire the product script of the first product and the product scripts of the n second products, where the product script consists of at least one link, and each link is configured with a link script; Retain the product presentation of the first product, merge the parts in the product presentations of the n second products that are repeated with the product presentation of the first product into the product presentation of the first product, and append the unmerged parts to the merged presentation to obtain the synthesized presentation for the recording task. Specifically, it includes: comparing each part of the product presentations of the n second products with each part of the product presentation of the first product in sequence, and finding the parts in the product presentations of the n second products that are repeated with the product presentation of the first product; merging the parts in the product presentations of the n second products that have the merge attribute and are repeated with the product presentation of the first product into the corresponding parts of the product presentation of the first product; appending the parts in the product presentations of the n second products that do not have the merge attribute and the parts that are not repeated with the product presentation of the first product to the merged presentation.

10. A storage medium, on which a computer program is stored, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

11. A computer device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

Images