Service recommendation method, system and device, server and storage medium

Abstract

The invention relates to the technical field of big data processing, in particular to a service recommendation method, system and device, a server and a storage medium. The method comprises the steps that when an incoming call request of a target user is received, a user portrait of the target user is called, and the user portrait comprises historical behavior information, basic information, historical service information and historical business information; the service labels are analyzed based on the user portraits, a correlation sequence is obtained, the correlation sequence comprises the correlation between the target user and each service label, and the correlation represents the probability that the user needs the service corresponding to the service label; determining a recommendation service type for the target user from the correlation sequence, and generating recommendation voice based on the recommendation service type; and then an incoming call request is accessed, and recommended voice is broadcasted for the target user. By adopting the method, the customer service efficiency can be improved.

Description

Technical Field

[0001] The present application relates to the technical field of big data processing, and in particular to a service recommendation method, system, device, server and storage medium. Background Art

[0002] In life, some service scenarios may be realized through voice calls, such as a telephone customer service platform of a certain platform. Users dial the corresponding customer service hotline to access the customer platform, and the customer service platform actively broadcasts relevant service templates to users, guides users to services, and finally connects users to corresponding service modules based on their choices.

[0003] However, in the related art, after the user accesses the customer platform, a fixed service template is broadcasted to each user, and these service templates are nested in multiple layers. The user needs to go through multiple levels of selection and it takes a long time to enter the required service module, which leads to low service efficiency for customers.

[0004] Therefore, how to improve service efficiency for customers is an urgent problem that needs to be solved. Summary of the invention

[0005] Based on this, it is necessary to provide a service recommendation method, system, device, server and storage medium to improve the service efficiency for customers in response to the above technical problems.

[0006] In a first aspect, the present application provides a service recommendation method, comprising:

[0007] Upon receiving an incoming call request from a target user, a user profile of the target user is retrieved, the user profile including historical behavior information, basic information, historical service information, and historical business information;

[0008] Analyze each service tag based on the user portrait to obtain a correlation sequence, wherein the correlation sequence includes the correlation between the target user and each service tag, and the correlation represents the probability that the user needs the service corresponding to the service tag;

[0009] Determining a recommended service type for the target user from the correlation sequence, and generating a recommended voice based on the recommended service type;

[0010] The incoming call request is received, and the recommended voice is broadcast to the target user.

[0011] In one embodiment, analyzing each service tag based on the user portrait to obtain a correlation sequence includes:

[0012] The user portrait is input into a pre-built decision tree model to obtain a correlation sequence output by the decision tree model; the decision tree model includes a decision tree corresponding to each item of information in the user portrait, and the decision trees are connected in parallel or in series.

[0013] In one embodiment, determining the recommended service type for the target user from the correlation sequence includes:

[0014] Sorting the service tags in the correlation sequence based on the correlation to obtain the sorted correlation sequence;

[0015] Determine a recommended tag from the sorted correlation sequence, wherein the recommended tag is the first N service tags with the highest correlation, wherein N is not less than 1;

[0016] Determine that the service type corresponding to the recommended tag is the recommended service type.

[0017] In one embodiment, generating a recommendation voice based on the recommended service type includes:

[0018] The recommended service and at least one of the basic information and the historical business information determine the service content, and match keywords corresponding to the service content from a preset dictionary;

[0019] Determine a speech template based on the recommended service type;

[0020] The keywords are filled into the speech template to obtain recommended speech, and the recommended speech is used to generate recommended voice.

[0021] In one embodiment, generating a recommendation voice based on the recommended service type includes:

[0022] The basic information is matched with preset personalized rules to obtain a tag set for the target user, wherein the personalized rules include a corresponding relationship between each item of the basic information and a tag; the tag set includes at least one of a content tag and a voice tag; the content tag is used to modify the recommended speech, and the voice tag represents at least one of the timbre, speech speed and volume of the voice;

[0023] Modify the recommended words based on the content tag to obtain personalized words;

[0024] Based on the personalized speech and the voice tag, a recommended voice is generated.

[0025] In one embodiment, the method further comprises:

[0026] If it is determined that the target user is not a historical user, the incoming call request is accepted and a template voice is broadcast to the target user.

[0027] In a second aspect, the present application also provides a service recommendation system, the system comprising a data management unit, a user feature scoring unit, a dialogue management unit and a voice interaction unit, wherein:

[0028] A data management unit, configured to retrieve a user profile of a target user upon receiving an incoming call request from the target user, wherein the user profile includes historical behavior information, basic information, historical service information, and historical business information;

[0029] A user feature scoring unit, configured to analyze each service tag based on the user portrait to obtain a correlation sequence, wherein the correlation sequence includes a correlation between the target user and each service tag, and the correlation represents a probability that the user requires a service corresponding to the service tag;

[0030] a dialogue management unit, configured to determine a recommended service type for the target user from the correlation sequence, and generate a recommended voice based on the recommended service type;

[0031] A voice interaction unit is used to receive the incoming call request and broadcast the recommended voice to the target user.

[0032] In a third aspect, the present application further provides a service recommendation device, the device comprising a portrait retrieval module, an analysis module, a voice generation module and an access module, wherein:

[0033] A portrait retrieval module is used to retrieve a user portrait of a target user upon receiving an incoming call request from the target user, wherein the user portrait includes historical behavior information, basic information, historical service information, and historical business information;

[0034] An analysis module, configured to analyze each service tag based on the user portrait to obtain a correlation sequence, wherein the correlation sequence includes a correlation between the target user and each service tag, and the correlation represents a probability that the user requires a service corresponding to the service tag;

[0035] A voice generation module, used to determine the recommended service type for the target user from the correlation sequence, and generate a recommendation voice based on the recommended service type;

[0036] The access module is used to access the incoming call request and broadcast the recommended voice to the target user.

[0037] In a fourth aspect, the present application further provides a server, comprising a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0038] Upon receiving an incoming call request from a target user, a user profile of the target user is retrieved, the user profile including historical behavior information, basic information, historical service information, and historical business information;

[0039] Analyze each service tag based on the user portrait to obtain a correlation sequence, wherein the correlation sequence includes the correlation between the target user and each service tag, and the correlation represents the probability that the user needs the service corresponding to the service tag;

[0040] Determining a recommended service type for the target user from the correlation sequence, and generating a recommended voice based on the recommended service type;

[0041] The incoming call request is received, and the recommended voice is broadcast to the target user.

[0042] In a fifth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the following steps are implemented:

[0043] Upon receiving an incoming call request from a target user, a user profile of the target user is retrieved, the user profile including historical behavior information, basic information, historical service information, and historical business information;

[0044] Analyze each service tag based on the user portrait to obtain a correlation sequence, wherein the correlation sequence includes the correlation between the target user and each service tag, and the correlation represents the probability that the user needs the service corresponding to the service tag;

[0045] Determining a recommended service type for the target user from the correlation sequence, and generating a recommended voice based on the recommended service type;

[0046] The incoming call request is received, and the recommended voice is broadcast to the target user.

[0047] In a sixth aspect, the present application further provides a computer program product, including a computer program, which implements the following steps when executed by a processor:

[0048] Upon receiving an incoming call request from a target user, a user profile of the target user is retrieved, the user profile including historical behavior information, basic information, historical service information, and historical business information;

[0049] Analyze each service tag based on the user portrait to obtain a correlation sequence, wherein the correlation sequence includes the correlation between the target user and each service tag, and the correlation represents the probability that the user needs the service corresponding to the service tag;

[0050] Determining a recommended service type for the target user from the correlation sequence, and generating a recommended voice based on the recommended service type;

[0051] The incoming call request is received, and the recommended voice is broadcast to the target user.

[0052] The above-mentioned service recommendation method, system, device, server and storage medium, when receiving an incoming call request from a target user, can retrieve the user portrait of the target user and conduct a comprehensive analysis of the user, thereby determining the correlation sequence of the target user, that is, determining the correlation between the target user and each of the service tags, and the correlation represents the probability that the user requires the service corresponding to the service tag; that is, by analyzing the user portrait of the target user, it is possible to predict the probability that the user requires the service corresponding to each service tag, and then, determine the recommended service type for the target user this time from the obtained correlation sequence, and the recommended service type is also the service tag with the highest predicted target user demand probability; then, a recommended voice is further generated for the recommended service type, and then, when receiving the incoming call request from the target user, the recommended voice is broadcast to the target user. Since the recommended service is predetermined, if the predicted recommended service type can hit the service that the target user needs to consult, it can reduce the waiting time of the user, thereby improving the efficiency of serving the target user. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the related technologies, the drawings required for use in the embodiments or the related technical descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0054] FIG1 is an application environment diagram of a service recommendation method in one embodiment;

[0055] FIG2 is a schematic diagram of a flow chart of a service recommendation method in one embodiment;

[0056] FIG3 is a schematic diagram of a process for determining a recommended service type in one embodiment;

[0057] FIG4 is a schematic diagram of a process of generating a recommended voice based on a recommended service type in one embodiment;

[0058] FIG5 is a schematic diagram of a process of generating a recommended speech based on a tag set in one embodiment;

[0059] FIG6 is a schematic diagram of the structure of a service recommendation system in one embodiment;

[0060] FIG7 is a structural block diagram of a service recommendation device in one embodiment;

[0061] FIG. 8 is a diagram showing the internal structure of a server in one embodiment. DETAILED DESCRIPTION

[0062] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0063] The service recommendation method provided in the embodiment of the present application can be applied in the application environment as shown in Figure 1. Among them, the terminal 102 communicates with the server 104 through the network; that is, the user can initiate an incoming service request to the server through the terminal 102, and after the server 104 receives the user's incoming request, it broadcasts the relevant service voice or relevant service data to the user. The data storage system can store the data that the server 104 needs to process. The data storage system can be integrated on the server 104, or it can be placed on the cloud or other network servers. Among them, the terminal 102 can be, but is not limited to, various personal computers, laptops, smart phones, tablet computers, Internet of Things devices and portable wearable devices. The Internet of Things devices can be smart speakers, smart TVs, smart air conditioners, smart car-mounted devices, etc. Portable wearable devices can be smart watches, smart bracelets, head-mounted devices, etc. The server 104 can be implemented with an independent server or a server cluster consisting of multiple servers.

[0064] In an exemplary embodiment, as shown in FIG. 2 , a service recommendation method is provided, which is applied to the server in FIG. 1 , and is described by taking the insurance business platform as the target platform, and includes the following steps 10 to 40, wherein:

[0065] Step 10: When receiving an incoming call request from a target user, retrieve the user profile of the target user, which includes historical behavior information, basic information, historical service information, and historical business information.

[0066] For the embodiment of the present application, the target user can be a new user or a historical user. The target user initiates an incoming call request to the server through the terminal. When the server receives the incoming call request from the target user, it obtains the user's identity from the accompanying information following the incoming call request, where the identity representation can be a mobile phone number, landline number or ID number, or other data that can represent the user's identity. The specific data type of the identity identifier is not limited in the embodiment of the present application, as long as the identity of the target user can be identified.

[0067] Specifically, the user portrait includes at least historical behavior information, basic information, historical service information and historical business information. Of course, based on actual business needs and scenarios, the user portrait may also include other types of information. Among them, historical behavior information includes the behavior data generated by the user on the APP, mini-program or web page related to the target platform; the basic information is the user's personal information entered by the user, such as the user's gender, occupation, age and place of residence; the basic information may also include the personal information of the user's family members entered by the user. Historical service information includes the types and content of services that the user has consulted on the target platform; historical business information includes the user's purchase orders, claims orders and cancellation data on the target platform. It should be noted that each item of information / data in the user portrait is associated with a corresponding generation time; the generation time represents the time when the information / data was last updated.

[0068] Step 20: Analyze each service tag based on the user portrait to obtain a correlation sequence, where the correlation sequence includes the correlation between the target user and each service tag, and the correlation represents the probability of the user requiring the service corresponding to the service tag.

[0069] For the embodiment of the present application, in the embodiment of the present application, the user portrait of each historical user exists in the data management unit. Based on the identity information of the target user obtained in step 10, it is determined whether the target user is a historical user. If the target user is a historical user, it indicates that the historical portrait of the user exists in the data management unit; if the target user is not a historical user, it indicates that the user is a new user, that is, the user portrait of the target user does not exist in the data management unit.

[0070] Furthermore, when it is determined that the target user is a historical user, the user portrait of the target user is queried and retrieved from the data management unit based on the identity identifier. The user portrait of the target user is analyzed to obtain a correlation sequence including the correlation between the target user and each service tag. In one possible implementation, the correlation sequence can be obtained by a pre-built analysis model. For example, an analysis model is obtained by training the user portraits of multiple users, and then the user portrait of the target user is output to the analysis model, and the correlation sequence for the target user output by the analysis model can be obtained. In another possible implementation, the correlation sequence can also be obtained by a preset algorithm or a pre-built analysis tool.

[0071] The service tag represents the type of service provided by the target platform; for example, the query service may correspond to the query tag 01, the order query service may correspond to the tag 011 corresponding to the order query, and the business progress query service may correspond to the tag 012 corresponding to the business query; the division method of the service tags is not specifically limited in the embodiments of the present application.

[0072] Step 30: Determine the recommended service type for the target user from the correlation sequence, and generate a recommended voice based on the recommended service type.

[0073] For the embodiments of the present application, the recommended service type is determined from the correlation sequence based on the preset recommendation rules; since the correlation represents the probability of the service corresponding to the user's required service tag, the preset recommendation rule can be to select the service tag with the greatest correlation as the recommended service tag; it can also be to determine the historical service type consulted by the target user last time based on the user's historical service information, and then, when the corresponding correlation of the historical service type is greater than the set threshold, the historical service type consulted by the target user last time is determined as the recommended service type for the target user this time. The specific logic of the recommendation rules is not specifically limited in the embodiments of the present application, as long as it is convenient to recommend service types that better meet the actual needs of users. Furthermore, the recommended service type can be one or more; the number of recommended types is not specifically limited in the embodiments of the present application.

[0074] After determining the recommended service type for the target user, a recommendation voice can be directly generated for the recommended service type; for example, if the recommended service type is query service, the query voice content can be "Excuse me, what information do you want to query?" After determining the recommended service type for the target user, it can also be determined whether there is data corresponding to the recommended service type in the user portrait, and a recommendation voice can be generated based on the data corresponding to the recommended service type; for example, if the recommended service type is order query, if the target user's historical business data shows that the user has an unfinished order A, the generated recommendation voice content can be "Excuse me, do you want to query the progress of order A?"

[0075] Step 40: Accept the incoming call request and announce the recommended voice to the target user.

[0076] In the embodiment of the present application, after generating the recommended voice for the target user, the incoming call request of the target user is received, and the recommended voice is immediately broadcasted to the target user when the incoming call request is received. Alternatively, when the incoming call request of the target user is received, a preset prompt voice is first broadcasted, and after the prompt voice broadcasting is finished, the recommended voice is broadcasted. The method of broadcasting the recommended voice is not specifically limited in the embodiment of the present application.

[0077] In the above service recommendation method, when an incoming call request from a target user is received, the user portrait of the target user can be retrieved to conduct a comprehensive analysis of the user, thereby determining the correlation sequence of the target user, that is, determining the correlation between the target user and each service tag, and the correlation represents the probability of the user requiring the service corresponding to the service tag; that is, by analyzing the user portrait of the target user, it is possible to predict the probability of obtaining the service corresponding to each service tag required by the user, and then, determine the recommended service type for the target user this time from the obtained correlation sequence, and the recommended service type is also the service tag with the highest predicted target user demand probability; then, a recommended voice is further generated for the recommended service type, and then, when receiving the incoming call request from the target user, the recommended voice is broadcast to the target user. Since the recommended service is predetermined, if the predicted recommended service type can hit the service that the target user needs to consult, it can reduce the user's waiting time, thereby improving the efficiency of serving the target user.

[0078] In one embodiment, in step 10, if it is determined that the target user is not a historical user, the incoming call request is immediately accepted and the template voice is broadcasted to the target user. That is to say, if the target user is not a historical user, then it is a new user, and there is no user portrait of the new user in the data management unit, so it is impossible to predict the service type required by the user. Therefore, the incoming call request is directly accepted and the template voice is broadcasted to the target user, so as to further obtain the information input by the target user, match the corresponding service type for the target user according to the information input by the target user, and generate the corresponding voice broadcast to broadcast to the target user.

[0079] Furthermore, in one of the embodiments, step 20 may specifically include: inputting the user portrait into a pre-built decision tree model to obtain a correlation sequence output by the decision tree model.

[0080] Specifically, in one possible implementation, the decision tree model includes a decision tree corresponding to each piece of information in the user portrait, and the decision trees are connected in parallel or in series.

[0081] In another possible implementation, multiple layers of nodes may be constructed in a decision tree, and each layer of nodes corresponds to a type of information in the user portrait. In the embodiment of the present application, a decision tree with multiple layers of nodes is constructed as an example for detailed description.

[0082] In an example:

[0083] For the root node of the decision tree: the feature selection is the basic information of the user (such as age, gender, place of residence, etc.); then the root node is branched according to the different attributes of the basic information, such as by age group, gender, etc.

[0084] Furthermore, for the first-level nodes of the decision tree: the feature selection is the user's historical behavior information (such as browsing history, click behavior, search history, etc.); branching is performed according to the attributes of the user's historical behavior, such as commonly used service types, most frequently visited pages, etc.

[0085] For the second-level nodes of the decision tree: the feature selection is the user's historical business information (such as purchase orders, refund orders, evaluation records, etc.); segmentation is performed based on the attributes of the user's business behavior, such as the category of purchased goods, purchase frequency, etc.

[0086] For the third-level nodes of the decision tree: the feature selection is the user's historical service information (such as customer support records, complaint records, service consultation records, etc.); segmentation is performed based on the user's service consultation situation, such as commonly used service tags, service feedback, etc.

[0087] For the leaf nodes of the decision tree: the output is a correlation sequence; each leaf node represents a specific service label and is accompanied by a matching score, indicating the degree of match between the user profile and the service label.

[0088] The above content is the structure of a decision tree constructed in an embodiment of the present application that can analyze user portraits; the structure of the above decision tree, the setting order of nodes in each layer, the feature selection method and the branching method are only an example of possible implementation, and are not a detailed limitation of the decision tree in the application scenario of the present application.

[0089] Further, in one embodiment, referring to FIG. 3 , the process of determining the recommended service type in step 30 may specifically include steps 31 to 33, wherein:

[0090] Step 31: sort the service tags in the correlation sequence based on the correlation to obtain a sorted correlation sequence;

[0091] Step 32: Determine a recommended tag from the sorted correlation sequence, where the recommended tag is the top N service tags with the highest correlation, where N is not less than 1;

[0092] Step 33: Determine that the service type corresponding to the recommended tag is the recommended service type.

[0093] Specifically, the service tags are sorted in descending order of relevance to obtain a sorted relevance sequence, and then the first N service tags are selected from the sorted relevance sequence, and then the selected N service tags are determined as recommended service types for the target user.

[0094] Among them, in one possible implementation, N can be a fixed value, for example, N can be set to 1 or 2. In another possible implementation, N can also be determined based on a certain type of information in the user profile; for example, N is determined based on the number of unfinished business orders in the historical business information, that is, it is determined that the target user has several unfinished business orders, and then the value of N is determined; of course, in actual application scenarios, N can also be determined based on other types of information in the user profile of the target user, which are not listed one by one in the embodiments of the present application.

[0095] Further, in one embodiment, referring to FIG. 4 , in step 30, the process of generating a recommended voice based on the recommended service type may specifically include steps 34 to 36, wherein:

[0096] Step 34: recommend a service and at least one of the basic information and historical business information, determine the service content, and match keywords corresponding to the service content from a preset dictionary.

[0097] Specifically, the recommended service type determines the service type. Different service types correspond to different service contents. For example, the service content of the order inquiry service needs to include the order number, while the service content of the manual inquiry service only needs to be transferred to a manual person. In an embodiment of the present application, the attributes of the keywords corresponding to each service type and the priority of each attribute can be preset; the attribute of the keyword is the type of parameter in the user's basic information and historical business information. After determining the recommended service type for the target user, a match is performed from the target user's basic information and / or historical business information to determine the keyword. For example, when it is determined that the recommended service type is a progress inquiry service, the order number of the unfinished order is determined as a keyword from the target user's historical business information.

[0098] Step 35: Determine a speech template based on the recommended service type;

[0099] Step 36: Fill the keywords into the speech template to obtain the recommended speech, and generate the recommended speech into a recommended voice.

[0100] Furthermore, matching speech templates are pre-set for different service types so that users can understand the service content more clearly. Specifically, each speech template includes a text structure and voice parameters, and the voice parameters may include at least one of the tone type, volume, timbre, and speech speed. For example, the text structure of the speech template corresponding to the progress inquiry service can be "Dear user, do you want to inquire about the progress of XX order?"; the expected type is a request type. For different service types, the corresponding speech templates are not exemplified one by one in this embodiment.

[0101] Furthermore, after determining the speech template corresponding to the recommended service type, fill the keywords determined in step 34 into the corresponding positions in the text structure of the speech template to obtain the recommended speech; further, the tone type of the speech template and the recommended speech are input into a pre-trained TTS (Text-to-Speech) model to obtain the recommended voice corresponding to the recommended speech.

[0102] Further, referring to FIG. 5 , in step 36 , the process of generating a recommendation voice based on the recommendation speech may specifically include steps 361 to 363 , wherein:

[0103] Step 361: Match the basic information from the preset personalization rules to obtain a tag set for the target user, wherein the personalization rules include the correspondence between each basic information and the tag; the tag set includes at least one of a content tag and a voice tag; the content tag is used to modify the recommended words, and the voice tag represents at least one of the timbre, speaking speed and volume of the voice.

[0104] Specifically, some data in the user's basic information can characterize the user's characteristics or preferences, and generating a recommended voice target user broadcast based on the user's characteristics or preferences can enhance the target user's perceived experience. In one example, the user's basic information shows that the user is 61 years old, so the volume of the recommended voice broadcast for the user should be slightly louder.

[0105] Step 362: modify the recommended words based on the content tag to obtain personalized words;

[0106] Step 363: Generate recommended voice based on personalized speech and voice tags.

[0107] Specifically, the correspondence between the data in the user's basic information and the label can be pre-set. Based on the set correspondence between the data and the label, the label set corresponding to the target user's basic information is determined, and the voice parameters in the recommended speech are further modified based on the label set to obtain the modified voice label.

[0108] Based on the modified voice tags and the determined recommendation words, the corresponding recommendation voice is generated, and the generated recommendation voice is broadcast to the target user.

[0109] It should be understood that, although the steps in the flowcharts involved in the above embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps is not strictly limited in order, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above embodiments may include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0110] In an exemplary embodiment, the embodiment of the present application provides a service recommendation system for implementing the service recommendation method involved above; as shown in FIG6 , the service recommendation system includes a data management unit, a user feature scoring unit, a dialogue management unit, and a voice interaction unit, wherein:

[0111] A data management unit, configured to retrieve a user profile of a target user upon receiving an incoming call request from the target user, wherein the user profile includes historical behavior information, basic information, historical service information, and historical business information;

[0112] A user feature scoring unit is used to analyze each service tag based on the user portrait to obtain a correlation sequence, which includes the correlation between the target user and each service tag. The correlation represents the probability of the service corresponding to the user's required service tag;

[0113] A dialogue management unit, used to determine a recommended service type for a target user from the correlation sequence, and generate a recommended voice based on the recommended service type;

[0114] The voice interaction unit is used to receive incoming call requests and broadcast recommended voice messages to target users.

[0115] As the entry point for user access, the voice interaction unit has key functions such as receiving user voice, voice recognition (ASR, Automatic Speech Recognition) and text-to-speech synthesis (TTS). It receives user voice input through the IVR (Interactive Voice Response) system, converts voice signals into text data, and then uses natural language processing (NLP, Natural Language Processing) technology for semantic recognition to determine the user's intention. At the same time, when the voice interaction unit plays the script to the user, it uses the TTS model to convert the text content into a voice stream to ensure that the broadcast is natural and smooth.

[0116] The user's incoming call information (such as incoming call number, call time, etc.) and the recognized text content will be sent to the data management unit and the user feature scoring unit. After receiving the recommended content (recommended words) returned by the dialogue management unit, the voice interaction unit will play the corresponding voice stream to the user.

[0117] The data management unit is the data center of the system, focusing on managing the multi-dimensional data of historical users, including voice navigation data, insurance policy data, APP interaction data, etc. This unit is responsible for integrating user information from various data sources to build a detailed user portrait and provide the latest user data and historical records every time the user calls in. After receiving the basic data of the user transmitted by the voice interaction unit, the data management unit sends the associated multi-dimensional user data to the user feature scoring unit for label scoring and feature analysis. Once the user information in the data management unit is updated, it will be synchronized to subsequent interactions to achieve dynamic optimization of user portraits.

[0118] The user feature scoring unit is the analysis and scoring center in the system, which mainly uses the random forest algorithm or decision tree model to score user tags. By calling the user data in the data management unit, the unit analyzes the user's call behavior, service preferences, consumption characteristics, etc. to calculate the user's tag score. The scoring results are used to guide the dialogue management unit to formulate exclusive service content and dialogue strategies. After obtaining the multi-dimensional data from the data management unit, the user feature scoring unit uses the random forest algorithm to score the tags and pass the scoring results to the dialogue management unit. In this way, the system can identify the user's demand level and service preference based on the score.

[0119] Furthermore, the service recommendation system of this embodiment also includes a dialogue management unit, which is the core of personalized interaction with users. The unit receives the user's tag score and matches the most suitable exclusive service for the user based on the user's score and preference information. The unit generates personalized dialogue content and mounts the corresponding voice script content so that it can be played to the user in the voice interaction unit. After obtaining the scoring data of the user feature scoring unit, the dialogue management unit matches the appropriate service content and returns the generated personalized script content to the voice interaction unit to realize voice-based active recommendation services and service guidance. The system's interaction records and user feedback will be synchronized back to the data management unit to help improve the user portrait.

[0120] The system of the present application combines the route data of incoming calls with user portraits, and provides proactive personalized services in voice navigation by extracting user features, so that users can obtain the required services more quickly and intuitively, reducing the number of operation steps and waiting time, and improving user satisfaction. On the other hand, traditional IVR systems often provide unified service options and cannot be personalized according to the specific needs and preferences of users. The data management unit establishes a detailed user portrait by saving and managing multi-dimensional user information. The service recommendation system of the embodiment of the present application can provide users with personalized service recommendations based on their user portraits and preferences. This ensures that the system has a more comprehensive understanding of the user and can provide a customized experience for each user.

[0121] Based on the same inventive concept, the embodiment of the present application also provides a service recommendation device for implementing the service recommendation method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more service recommendation device embodiments provided below can refer to the limitations on the service recommendation method above, and will not be repeated here.

[0122] In an exemplary embodiment, as shown in FIG. 7 , a service recommendation device is provided, including a portrait retrieval module 701, an analysis module 702, a voice generation module 703, and an access module 704, wherein:

[0123] The portrait retrieval module 701 is used to retrieve the user portrait of the target user when receiving an incoming call request from the target user. The user portrait includes historical behavior information, basic information, historical service information, and historical business information;

[0124] An analysis module 702 is used to analyze each service tag based on the user portrait to obtain a correlation sequence, where the correlation sequence includes the correlation between the target user and each service tag, and the correlation represents the probability of the service corresponding to the service tag required by the user;

[0125] A voice generation module 703, used to determine the recommended service type for the target user from the correlation sequence, and generate a recommended voice based on the recommended service type;

[0126] The access module 704 is used to receive the incoming call request and broadcast the recommended voice to the target user.

[0127] In the above-mentioned service recommendation device, when receiving an incoming call request from a target user, the user portrait of the target user can be retrieved to perform a comprehensive analysis on the user, thereby determining the correlation sequence of the target user, that is, determining the correlation between the target user and each service tag, and the correlation represents the probability of the user requiring the service corresponding to the service tag; that is, by analyzing the user portrait of the target user, it is possible to predict the probability of the user requiring the service corresponding to each service tag, and then, determine the recommended service type for the target user this time from the obtained correlation sequence, and the recommended service type is also the service tag with the highest predicted target user demand probability; then, a recommended voice is further generated for the recommended service type, and then, when receiving the incoming call request from the target user, the recommended voice is broadcast to the target user. Since the recommended service is predetermined, if the predicted recommended service type can hit the service that the target user needs to consult, it can reduce the user's waiting time, thereby improving the efficiency of serving the target user.

[0128] In one embodiment, the analysis module 702 is specifically configured to:

[0129] The user profile is input into a pre-built decision tree model to obtain a correlation sequence output by the decision tree model; the decision tree model includes a decision tree corresponding to each piece of information in the user profile, and the decision trees are connected in parallel or in series.

[0130] In one embodiment, the speech generation module 703 is specifically used for:

[0131] Sorting the service labels in the correlation sequence based on the correlation to obtain a sorted correlation sequence;

[0132] Determine a recommended tag from the sorted correlation sequence, where the recommended tag is the top N service tags with the highest correlation, where N is not less than 1;

[0133] Determine that the service type corresponding to the recommended label is the recommended service type.

[0134] In one embodiment, the speech generation module 703 is specifically used for:

[0135] Recommending a service and at least one of the basic information and historical business information, determining the service content, and matching keywords corresponding to the service content from a preset dictionary;

[0136] Determine the script template based on the recommended service type;

[0137] Fill the keywords into the speech template to obtain the recommended speech, and generate the recommended speech into the recommended voice.

[0138] In one embodiment, the speech generation module 703 is specifically used for:

[0139] The basic information is matched from the preset personalized rules to obtain a tag set for the target user, wherein the personalized rules include the correspondence between each basic information and the tag; the tag set includes at least one of a content tag and a voice tag; the content tag is used to modify the recommended speech, and the voice tag represents at least one of the timbre, speech speed and volume of the voice;

[0140] Modify the recommended words based on content tags to obtain personalized words;

[0141] Generate recommended voice based on personalized words and voice tags.

[0142] In one embodiment, the service recommendation device further includes a template broadcast module, and the template broadcast module is specifically used to:

[0143] If it is determined that the target user is not a historical user, the incoming call request is accepted and the template voice is broadcast to the target user.

[0144] Each module in the above service recommendation device can be implemented in whole or in part by software, hardware or a combination thereof. Each module can be embedded in or independent of a processor in a server in the form of hardware, or can be stored in a memory in a server in the form of software, so that the processor can call and execute the corresponding operations of each module.

[0145] In an exemplary embodiment, a server is provided, which may be a terminal, and its internal structure diagram may be shown in FIG8. The server includes a processor, a memory, an input / output interface, a communication interface, a display unit, and an input device. The processor, the memory, and the input / output interface are connected via a system bus, and the communication interface, the display unit, and the input device are connected to the system bus via the input / output interface. The processor of the server is used to provide computing and control capabilities. The memory of the server includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the server is used to exchange information between the processor and an external device. The communication interface of the server is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner may be implemented through WIFI, a mobile cellular network, NFC (near field communication) or other technologies. When the computer program is executed by the processor, a service recommendation method is implemented. The display unit of the server is used to form a visually visible picture, which may be a display screen, a projection device, or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen, and the input device of the server can be a touch layer covered on the display screen, or a button, trackball or touchpad set on the server shell, or an external keyboard, touchpad or mouse.

[0146] Those skilled in the art will understand that the structure shown in FIG. 8 is merely a block diagram of a partial structure related to the solution of the present application, and does not constitute a limitation on the server to which the solution of the present application is applied. A specific server may include more or fewer components than shown in the figure, or combine certain components, or have a different arrangement of components.

[0147] In an exemplary embodiment, a server is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the steps of the service recommendation method described in any one of the above method embodiments are implemented.

[0148] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of the service recommendation method described in any one of the above method embodiments are implemented.

[0149] In one embodiment, when the computer program is executed by a processor, it also implements the steps of the service recommendation method as described in any one of the above method embodiments.

[0150] In one embodiment, a computer program product is provided, including a computer program, which, when executed by a processor, implements the steps of the service recommendation method as described in any one of the above method embodiments.

[0151] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0152] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.

[0153] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0154] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A service recommendation method, characterized in that: The method comprises: Upon receiving an incoming call request from a target user, a user profile of the target user is retrieved, the user profile including historical behavior information, basic information, historical service information, and historical business information; Analyze each service tag based on the user portrait to obtain a correlation sequence, wherein the correlation sequence includes the correlation between the target user and each service tag, and the correlation represents the probability that the user needs the service corresponding to the service tag; Determining a recommended service type for the target user from the correlation sequence, and generating a recommended voice based on the recommended service type; The incoming call request is received, and the recommended voice is broadcast to the target user.

2. The method according to claim 1, characterized in that The analyzing each service tag based on the user portrait to obtain a correlation sequence includes: The user portrait is input into a pre-built decision tree model to obtain a correlation sequence output by the decision tree model; the decision tree model includes a decision tree corresponding to each item of information in the user portrait, and the decision trees are connected in parallel or in series.

3. The method according to claim 1, characterized in that The determining the recommended service type for the target user from the correlation sequence includes: Sorting the service tags in the correlation sequence based on the correlation to obtain the sorted correlation sequence; Determine a recommended tag from the sorted correlation sequence, wherein the recommended tag is the first N service tags with the highest correlation, wherein N is not less than 1; Determine that the service type corresponding to the recommended tag is the recommended service type.

4. The method according to any one of claims 1 to 3, characterized in that The generating of the recommended voice based on the recommended service type comprises: The recommended service and at least one of the basic information and the historical business information determine the service content, and match keywords corresponding to the service content from a preset dictionary; Determine a speech template based on the recommended service type; The keywords are filled into the speech template to obtain recommended speech, and the recommended speech is used to generate recommended voice.

5. The method according to claim 4, characterized in that The generating of the recommended voice based on the recommended service type comprises: The basic information is matched with preset personalized rules to obtain a tag set for the target user, wherein the personalized rules include a corresponding relationship between each item of the basic information and a tag; the tag set includes at least one of a content tag and a voice tag; the content tag is used to modify the recommended speech, and the voice tag represents at least one of the timbre, speech speed and volume of the voice; Modify the recommended words based on the content tag to obtain personalized words; Based on the personalized speech and the voice tag, a recommended voice is generated.

6. The method according to claim 1, characterized in that The method further comprises: If it is determined that the target user is not a historical user, the incoming call request is accepted and a template voice is broadcast to the target user.

7. A service recommendation system, characterized in that: The system includes a data management unit, a user feature scoring unit, a dialogue management unit and a voice interaction unit, wherein: A data management unit, configured to retrieve a user profile of a target user upon receiving an incoming call request from the target user, wherein the user profile includes historical behavior information, basic information, historical service information, and historical business information; A user feature scoring unit, configured to analyze each service tag based on the user portrait to obtain a correlation sequence, wherein the correlation sequence includes a correlation between the target user and each service tag, and the correlation represents a probability that the user requires a service corresponding to the service tag; a dialogue management unit, configured to determine a recommended service type for the target user from the correlation sequence, and generate a recommended voice based on the recommended service type; A voice interaction unit is used to receive the incoming call request and broadcast the recommended voice to the target user.

8. A service recommendation device, characterized in that: The device includes a portrait retrieval module, an analysis module, a voice generation module and an access module, wherein: A portrait retrieval module is used to retrieve a user portrait of a target user upon receiving an incoming call request from the target user, wherein the user portrait includes historical behavior information, basic information, historical service information, and historical business information; An analysis module, configured to analyze each service tag based on the user portrait to obtain a correlation sequence, wherein the correlation sequence includes a correlation between the target user and each service tag, and the correlation represents a probability that the user requires a service corresponding to the service tag; A voice generation module, used to determine the recommended service type for the target user from the correlation sequence, and generate a recommendation voice based on the recommended service type; The access module is used to access the incoming call request and broadcast the recommended voice to the target user.

9. A server, comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

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 method according to any one of claims 1 to 6 are implemented.

Images