Interactive service reservation method and device, computer equipment and readable storage medium

Abstract

The invention discloses an interactive service reservation method and device, computer equipment and a readable storage medium, relates to the technical field of computer software, can be applied to the field of medical health, constructs a voice interaction-intelligent matching-automatic guiding service process, and realizes convenience, precision and high efficiency of service reservation. The method comprises the steps of collecting a session audio input by a user, determining a user intention corresponding to the session audio, determining a user voiceprint corresponding to the session audio based on a voiceprint recognition technology when an intention type of the user intention is matched with a preset intention type, and logging in a target service platform according to the user voiceprint; when the intention type is a reservation intention, determining a to-be-reserved service and reservation information, querying a reservation standard corresponding to the to-be-reserved service, and comparing the reservation information with the reservation standard; and if the comparison result indicates that the reservation information is matched with the reservation standard, docking a service institution providing the to-be-reserved service based on the target service platform, and sending the reservation information to the service institution.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer software, which can be applied to the field of medical health, and in particular relates to an interactive service reservation method and device, a computer device and a readable storage medium. BACKGROUND

[0002] With the continuous development of the field of digital medicine, users' demand for online reservation business services is increasing, such as medical examination reservation scenarios. In related technologies, online reservation services mainly rely on APP or website platforms and are realized through graphical interface interaction and supplemented by manual services. However, the traditional online reservation method has a complex graphical interface design, which is extremely unfriendly to the elderly and non-technical users, has a high operation threshold, and leads to a high reservation failure rate. In addition, during the peak period of business, the server load pressure increases dramatically, and congestion problems frequently occur, so that users not only face the risk of reservation failure but also have to experience long reservation queuing. Thus, the existing online reservation technology has the technical problem of low reservation efficiency in the field of medical health. SUMMARY

[0003] Therefore, the present application provides an interactive service reservation method and device, a computer device and a readable storage medium, which mainly aims to solve the problem that the traditional online reservation method has a complex graphical interface design, which is extremely unfriendly to the elderly and non-technical users, has a high operation threshold, and leads to a high reservation failure rate. In addition, during the peak period of business, the server load pressure increases dramatically, and congestion problems frequently occur, so that users not only face the risk of reservation failure but also have to experience long reservation queuing. Thus, the existing online reservation technology has the technical problem of low reservation efficiency in the field of medical health.

[0004] According to a first aspect of the present application, an interactive service reservation method is provided, which comprises the following steps:

[0005] Collecting conversation audio input by a user, determining a user intent corresponding to the conversation audio, and based on a voiceprint recognition technology, determining a user voiceprint corresponding to the conversation audio when an intent type of the user intent matches a preset intent type, logging into a target service platform according to the user voiceprint, and the target service platform is used to purchase and reserve business services;

[0006] When the intent type is a reservation intent, determining a service to be reserved and reservation information according to the conversation audio, and querying a reservation standard corresponding to the service to be reserved, and comparing the reservation information with the reservation standard;

[0007] If the comparison result indicates that the reservation information matches the reservation standard, a service institution providing the to-be-reserved service based on the target service platform docking is determined, the reservation information is sent to the service institution, and after receiving reservation success information returned by the service institution, the reservation success information is fed back through the smart sound equipment and / or a specified channel.

[0008] According to a second aspect of the present application, an interactive service reservation device is provided, which comprises:

[0009] The acquisition module is configured to acquire conversation audio input by a user, determine a user intention corresponding to the conversation audio, and when an intention type of the user intention matches a preset intention type, determine a user voiceprint corresponding to the conversation audio based on voiceprint recognition technology, and log in a target service platform according to the user voiceprint, the target service platform being configured to purchase and reserve a business service.

[0010] The query module is configured to, when the intention type is a reservation intention, determine a to-be-reserved service and reservation information according to the conversation audio, and query a reservation standard corresponding to the to-be-reserved service, and compare the reservation information with the reservation standard.

[0011] The reservation module is configured to, if the comparison result indicates that the reservation information matches the reservation standard, send the reservation information to a service institution providing the to-be-reserved service based on the target service platform docking, and after receiving reservation success information returned by the service institution, feed back the reservation success information through the smart sound equipment and / or a specified channel.

[0012] According to a third aspect of the present application, a computer device is provided, which comprises a memory and a processor, the memory stores a computer program, and the processor implements the steps of the method of any one of the first aspect when executing the computer program.

[0013] According to a fourth aspect of the present application, a computer readable storage medium is provided, which stores a computer program, and the computer program implements the steps of the method of any one of the first aspect when executed by a processor.

[0014] By means of the technical scheme, the application provides an interactive service reservation method and device, computer equipment and a readable storage medium. The embodiment of the application integrates voiceprint recognition, natural language processing (NLP) and multi-modal interaction technology on a smart speaker device, constructs an interactive service reservation system, replaces the graphical interface operation of a traditional APP / website, and reduces the operation threshold of a user. In terms of technical application, the processes of voiceprint automatic login, information intelligent comparison and multi-round interactive guidance are automatically executed by the system, human intervention is minimized, service reservation efficiency is greatly improved, and the problem of reservation failure caused by operation errors is effectively avoided. The system finally constructs a closed-loop service process of "voice interaction-intelligent matching-automatic guidance", and realizes the convenience, accuracy and efficiency of service reservation.

[0015] The above description is only a summary of the technical scheme of the application. In order to more clearly understand the technical means of the application, the application can be implemented according to the content of the specification, and in order to make the above and other purposes, characteristics and advantages of the application more obvious and easy to understand, the following specific embodiments of the application are described. BRIEF DESCRIPTION OF DRAWINGS

[0016] Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The accompanying drawings are included to provide a description of the preferred embodiments and are not meant to limit the scope of the present application. Moreover, the same reference numerals in the attached drawings indicate the same or similar components. In the drawings:

[0017] Figure 1 A flowchart of an interactive service reservation method provided by an embodiment of the application is shown;

[0018] Figure 2 is Figure 1 A flowchart of a specific embodiment of step S10 in the method is shown;

[0019] Figure 3 A flowchart of another interactive service reservation method provided by an embodiment of the application is shown;

[0020] Figure 4 A flowchart of an interactive guide user to select a reservation item provided by an embodiment of the application is shown;

[0021] Figure 5 is Figure 1 A flowchart of a specific embodiment of step S30 in the method is shown;

[0022] Figure 6 A flowchart of another interactive user session recognition method provided by an embodiment of the application is shown;

[0023] Figure 7Another interactive service reservation method provided by the embodiment of the present application is shown in a flowchart;

[0024] Figure 8 Another interactive service reservation method provided by the embodiment of the present application is shown in a flowchart;

[0025] Figure 9 Another interactive service reservation method provided by the embodiment of the present application is shown in a flowchart;

[0026] Figure 10 Another interactive service reservation method provided by the embodiment of the present application is shown in a flowchart; DETAILED DESCRIPTION

[0027] The embodiments of the present application are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar notations represent the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by referring to the drawings are exemplary only, and are used only for explaining the present application, and cannot be interpreted as limiting the present application.

[0028] Those skilled in the art can understand that, unless specifically stated otherwise, the singular forms "one", "a", "said" and "the" as used herein include plural forms. It should be further understood that the phrase "comprising" used in the specification of the present application means that the features, integers, steps, operations, elements and / or components exist, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof.

[0029] Those skilled in the art can understand that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as that understood by those skilled in the art. It should also be understood that terms such as those defined in a general dictionary should be understood to have meanings consistent with those in the context of the prior art, and should not be interpreted in an idealized or overly formal sense unless specifically defined as such.

[0030] Those skilled in the art of the technology will understand that the "terminal" used herein includes not only a device with only a wireless signal receiver, but also a device with receiving and transmitting hardware capable of performing bidirectional communication on a bidirectional communication link. Such devices can include cellular or other communication devices with single-line or multi-line displays, or without multi-line displays; PCS (Personal Communications Service) devices that can combine voice, data processing, fax, and / or data communication capabilities; PDAs (Personal Digital Assistants) that can include radio frequency receivers, pagers, Internet / Intranet access, Web browsers, organizers, calendars, and / or GPS (Global Positioning System) receivers; conventional laptop and / or palmtop computers, or other devices that have and / or include radio frequency receivers. The "terminal" used herein can be portable, transportable, installed in vehicles (air, sea, and / or land), or suitable and / or configured for local operation, and / or distributed operation in any other location on Earth and / or in space. The "terminal" used herein can also be a communication terminal, an Internet terminal, a music / video playing terminal, such as a PDA, a MID (Mobile Internet Device), and / or a mobile phone with music / video playing functions, a smart TV, a set-top box, and the like.

[0031] The interactive service reservation method provided by the embodiments of the present application is suitable for a service reservation system of a smart sound box device. In actual operation, the service reservation system can be a physical examination reservation service system, and the present application takes the physical examination reservation service system as an example for illustration.

[0032] A smart sound box is a smart terminal device taking voice interaction as the core, integrating artificial intelligence, Internet of Things, cloud computing, and the like. The smart sound box usually has the audio playing function of a sound box and can realize various intelligent services through voice instructions. Without manual operation, the instructions can be completed through voice, which is suitable for users of all ages, especially convenient for the elderly and children.

[0033] The embodiments of the present application provide an interactive service reservation method, as shown in Figure 1 The method comprises the following steps.

[0034] S10, collect the conversation audio input by the user, determine the user intent corresponding to the conversation audio, and when the intent type of the user intent matches the preset intent type, determine the user voiceprint corresponding to the conversation audio based on voiceprint recognition technology, log in to a target service platform according to the user voiceprint, and the target service platform is used for purchasing and reserving business services.

[0035] In the embodiments of the present application, as shown in Figure 2 The intelligent sound box device continuously collects the conversation audio input by the user by performing the following steps S11 to S14, processes the audio through voiceprint recognition technology, and extracts and determines the corresponding user voiceprint features. Further, the system automatically associates and logs in to a physical examination reservation platform, i.e., a target service platform, which is used to support the user to complete the purchase and reservation operation of business services.

[0036] It can be understood that in the process of continuously collecting audio based on the conversation context, if the system detects that the user's expression has semantic ambiguity or receives a multi-modal input instruction, a multi-round interactive clarification process will be automatically started. For example, if the received voice instruction conflicts with the touch screen operation instruction, the system will start a multi-round interactive clarification process. Through this process, the system will actively dialogue with the user to clarify the ambiguous content or solve the instruction conflict, ensuring the accuracy and fluency of the reservation operation.

[0037] S11, collect the user's voice signal through the microphone array configured by the intelligent sound box device, obtain the conversation audio, and convert the conversation audio into conversation text through automatic speech recognition (ASR) technology.

[0038] In this step, in order to realize a natural, smooth, efficient and accurate human-computer voice interaction system, a semantic interaction module is called, and the voice interaction module collects the voice signal of the user's speech in real time through the microphone array, such as a ring array or a linear array, built-in the intelligent sound box device, to form a continuous conversation audio stream. Further, the voice interaction module uses automatic speech recognition (ASR) technology to process the audio stream, converts the voice signal into corresponding text content through the collaborative operation of an acoustic model, a language model and a pronunciation dictionary, and obtains the conversation text.

[0039] S12, use an NLP model to perform semantic understanding on the conversation text, determine the user intent and the intent type corresponding to the user intent.

[0040] In this step, the system dynamically calls the NLP model of the corresponding field based on the user's appointment service type. For example, if the user needs to make an appointment for financial services, the NLP model of the financial field is used for semantic understanding of the conversation text. If the user needs to make an appointment for medical services, the NLP model of the medical health field is used for semantic understanding of the conversation text. Specifically, the system calls the semantic understanding module, converts the user's voice input into a text string through ASR technology, for example, converts "I want to make an appointment for a physical examination next Wednesday" into text. Further, delete the mood words, punctuation marks in the text, and unify the number format, such as converting "next Wednesday" to "June 18, 2025" and "early tomorrow at 10 o'clock" to "June 14, 2025 10:00". Next, use the corresponding field-specific tokenizer to perform text segmentation on the text conversation, for example, split "cardiac color Doppler ultrasound examination" into heart, color Doppler, and examination, and label examination as a verb and heart as an entity noun. Further, based on a deep learning model, an intent classifier is constructed, the text segmentation is input into the intent classifier, and the user's intent output by the intent classifier is obtained, for example, the input text "help me recommend a physical examination package suitable for a 40-year-old woman", and the model identifies it as "recommendation intent - package recommendation".

[0041] It can be understood that colloquial expressions may appear in the conversation text input by the user. In order to improve the accuracy of semantic understanding, the system can also introduce a medical health field knowledge base to correct the ambiguity of colloquial expressions through the knowledge base. For example, the system introduces a medical field knowledge base, when the user says "check the heart", it is mapped to specific items such as electrocardiogram and cardiac color Doppler ultrasound through the knowledge base, and combined with the context to ask further.

[0042] S13, if the intent type of the user intent matches the preset intent type, the voiceprint feature vector of the voice signal is extracted, the voiceprint feature vector is compared with the pre-stored login voiceprint in the voiceprint library, and the login voiceprint matching the voiceprint feature vector is determined.

[0043] In this step, when the NLP model identifies that the user's intent is of a preset type, the dynamic identity binding module is triggered, wherein the preset type includes an appointment intent and a recommendation intent. In actual operation, the preset type can be flexibly configured according to the actual application scenario, for example, extended to a purchase intent, a cancel appointment intent, etc., and an operator can set it according to actual needs. Further, the dynamic identity binding module can extract a voiceprint feature vector from the voice signal through a Mel frequency cepstrum coefficient (MFCC) or a deep neural network (DNN) feature extraction algorithm. Then, the voiceprint feature vector is compared with the pre-stored user login voiceprint in the voiceprint library through cosine similarity calculation or dynamic time warping (DTW) comparison. Further, the user's identity is quickly matched without the need for the user to manually input an account password, realizing "opening to login" and non-intrusive verification. It solves the cumbersome process of traditional identity verification that requires manual input of information, and improves the continuity of interaction.

[0044] S14, determine the login voiceprint associated account login information, according to the account login information, through the OAuth protocol login target service platform.

[0045] In this step, the associated account information in the user health record database is retrieved by the dynamic identity binding module, and the account information at least includes the user ID and the purchased benefit data. Further, the account information is transmitted to the target service platform, such as a physical examination reservation system, through a TLS1.3 encrypted channel by using the authorization mechanism of the OAuth protocol, and the automatic login and benefit retrieval are completed. The encryption process can use the AES-256 algorithm combined with a session key generated by a dynamic random number.

[0046] S20, when the user intent corresponding to the conversation audio is a reservation intent, determining the to-be-reserved service and the reservation information according to the conversation audio, and querying the reservation standard corresponding to the to-be-reserved service, and comparing the reservation information with the reservation standard.

[0047] In the embodiments of the present application, as shown in Figure 3 After the system logs in the target service platform, if the user intent of the user is a reservation intent, it is also necessary to determine whether the user account is associated with the corresponding to-be-reserved service. If the corresponding to-be-reserved service is not associated, the system also needs to guide the user to purchase the related reservation project corresponding to the to-be-reserved service.

[0048] For example, in the medical health field, the to-be-reserved service is a reservation physical examination service, at this time, it is necessary to determine whether there is a to-be-used physical examination product (reservation project) under the user account. If not, the user needs to be guided to purchase the corresponding physical examination product (reservation project), and then perform the related reservation operation after the purchase is completed. If there is, it is also necessary to determine the number of reservation projects corresponding to the to-be-reserved service associated with the user account, if there are multiple reservation projects, as shown in Figure 4 , it is also necessary to guide the user to select a specific reservation project based on the conversation context, and then as shown in Figure 5 , the following steps S21 to S23 are executed to query the reservation standard corresponding to the to-be-reserved service, compare the reservation information with the reservation standard, and then perform service reservation according to the comparison result.

[0049] S21, using an NLP model to perform semantic understanding on the conversation audio, identifying the user intent and entity information, the entity information including the to-be-reserved service and the reservation information, the reservation information including the reservation time, the reservation agency and the reservation project.

[0050] In this step, as shown in Figure 6As shown, the system adopts the NLP model of the corresponding field to perform semantic understanding on the conversation audio, identify the user intent and entity information, and the entity information includes the to-be-reserved service and reservation information. Specifically, the system captures the multi-turn conversation dependency through the Transformer architecture. If the user implicitly or directly mentions the time, institution, and project through "one-sentence reservation" or the like in the conversation, the corresponding time entity is directly extracted as the reservation time, the corresponding institution entity is extracted as the reservation institution, the corresponding project entity is extracted as the reservation project, and the category corresponding to the reservation project is taken as the to-be-reserved service. For example, the user inputs "help me reserve a job physical examination at 10 o'clock on Sunday, and select the 0001 job physical examination package of XX institution". Among them, the time "10 o'clock on Sunday", the institution "XX institution", and the project "0001 job physical examination package" are directly mentioned, and the category "job physical examination" corresponding to the "0001 job physical examination package" can be directly taken as the reservation service, the "0001 job physical examination package" is taken as the reservation project, the "XX institution" is taken as the reservation institution, and the "10 o'clock on Sunday" is taken as the reservation time. In addition, it needs to be explained that some physical examination projects in the package are not available for unmarried women. At this time, as shown in the figure, the one-sentence reservation process also needs to obtain the marriage status of the user as user feature data when the user reserving the project is a woman. Figure 7

[0051] It can be understood that if the user input conversation content lacks the to-be-reserved service, the system can call the voice interaction module to initiate an inquiry to the user until the user supplements the missing to-be-reserved service. Then, it is determined whether the target service platform provides the to-be-reserved service. For example, the user inputs "help me reserve a job physical examination", and the system needs to determine whether the medical institutions registered on the target service platform can provide job physical examination services. If the target service platform can provide the to-be-reserved service, and if the conversation content still lacks reservation elements, the voice interaction module is called again to initiate an inquiry to the user until the user supplements the missing reservation elements, and then the following step S22 is performed to query the to-be-reserved project corresponding to the to-be-reserved service, obtain the reservation standard according to the query result, and compare the reservation information with the reservation standard.

[0052] It can be understood that if the medical institutions registered on the target service platform cannot provide the to-be-reserved service required by the user, the following step S23 is performed to generate a reservation failure prompt indicating that the user has no right to reserve the related service, and the reservation failure prompt is fed back through voice and / or vision.

[0053] S22, if the target service platform provides the to-be-reserved service, the to-be-reserved project corresponding to the to-be-reserved service is queried, the reservation standard is obtained according to the query result, and the reservation information is compared with the reservation standard.

[0054] ​In this step, if the target service platform provides a service to be booked, such as an entry physical examination service, all the to-be-booked items corresponding to the to-be-booked service are queried. Further, it is judged whether there is an item matching the booking item in the to-be-booked items. If there are multiple to-be-booked items associated with the to-be-booked service and there is no booking item in the booking information, a multi-round interactive clarification process is triggered, and the user selects the booking item in all the to-be-booked items through the intelligent voice device to obtain the booking standard corresponding to the booking item, and the booking information is compared with the booking standard. If there are multiple to-be-booked items associated with the to-be-booked service and there is a booking item in the booking information, the booking standard corresponding to the booking item is obtained in the case that the booking item hits all the to-be-booked items, and the booking information is compared with the booking standard. If there is one to-be-booked item associated with the to-be-booked service and there is a booking item in the booking information, the booking standard corresponding to the booking item is obtained in the case that the to-be-booked item matches the booking item, and the booking information is compared with the booking standard.

[0055] For example, the physical examination reservation platform provides multiple entry physical examination packages (to-be-booked items), which are package A, package B and package C respectively. The number of physical examination items and the applicable institutions of each package are different. When the user does not select a specific package (booking item), the system will call the dialogue management module to trigger a multi-round interactive clarification process, and the user selects one package as the booking item from package A, package B and package C through the intelligent voice device. When the user selects package B, it means that the booking item hits all the to-be-booked items, and the booking standard corresponding to package B can be obtained directly. When the user inputs package D, it means that the booking item does not hit all the to-be-booked items, and a multi-round interactive clarification process will be triggered to determine whether to cancel the reservation or replace the booking item by the user.

[0056] Further, the system converts the booking standard into a structured matching rule library through NLP technology, and standardizes the booking information into structured data, including the booking item, the booking time and the booking institution. The historical feature data corresponding to the user in the user feature database is obtained. If the historical feature data matches the user attributes recorded in the booking standard, it is determined that the user attribute matching is successful, otherwise, it is determined that the user attribute standard item matching fails. For example, there are some physical examination items in package A that unmarried women cannot perform. At this time, if the user booking the item is a woman, the user's marital status needs to be obtained and added to the user feature data, and then all the collected user feature data is compared with the user attributes recorded in the booking standard. If the historical feature data matches the user attributes recorded in the booking standard, it is determined that the user attribute matching is successful, otherwise, it is determined that the user attribute standard item matching fails.

[0057] Meanwhile, if the appointment time hits the recordable time in the appointment standard, it is determined that the appointment time matches successfully, otherwise, it is determined that the appointment time standard item matches unsuccessfully. If the appointment institution hits the recordable institution in the appointment standard, it is determined that the appointment institution matches successfully, otherwise, it is determined that the appointment institution standard item matches unsuccessfully. All the recordable projects associated with the appointment institution in the appointment standard are queried, if the appointment project hits the recordable project, it is determined that the appointment project matches successfully, otherwise, it is determined that the appointment project standard item matches unsuccessfully.

[0058] Further, if the user attribute, the appointment project, the appointment time and the appointment institution all match successfully, a comparison result indicating that the appointment information matches the appointment standard is generated. If any standard item in the user attribute, the appointment project, the appointment time and the appointment institution matches unsuccessfully, a comparison result indicating that the appointment information does not match the appointment standard is generated.

[0059] S23, if the target service platform does not provide the service to be reserved, a reservation failure prompt indicating that the user has no right to reserve the related service is generated, and the reservation failure prompt is fed back through voice and / or visual feedback.

[0060] In the embodiments of the present application, in order to inform the user of the reservation situation in time, for the smart sound box device with a display screen, a voice broadcast combined with visual prompt is adopted to feed back the prompt information. For the device supporting only voice interaction, a preset speech template is adopted to generate detailed explanation feedback prompt information.

[0061] S30, if the comparison result indicates that the appointment information matches the appointment standard, the appointment information is sent to the service institution based on the target service platform docking the service institution providing the service to be reserved, and after receiving the reservation success information returned by the service institution, the reservation success information is fed back through the smart sound device and / or the specified channel.

[0062] In this step, if the comparison result indicates that the appointment information matches the appointment standard, a secure communication channel with the service institution providing the service to be reserved is established based on the service directory of the target service platform, such as Figure 8As shown, the reservation information confirmation interaction flow is generated according to the reservation information, and the user is prompted to confirm the reservation information again through multiple rounds of dialogue. After the user confirms that the information is correct, a reservation form is generated according to the reservation information provided by the user, and the reservation form is pushed to the information system of the service institution through a secure data channel. After receiving the reservation success information returned by the service institution, the reservation success information is fed back through the intelligent sound equipment and / or the specified channel. The specified channel can be a communication address provided by the user through which a prompt message is sent to the user by short message. It should be noted that once the user's modification intention is identified through semantic analysis, the variable parameters in the reservation information are dynamically adjusted, and the comparison process of step S20 is re-executed to verify whether the modified information meets the reservation standard.

[0063] It can be understood that if the comparison result indicates that the reservation information does not match the reservation standard, the unmatched reservation standard item is determined, a guided interaction session is generated according to the reservation standard item, and the guided interaction session is output through voice and / or vision. Based on the context of the session, the session audio input by the user is continuously collected, the target business service and the reservation information are updated, the interface with the resource institution, the reservation standard comparison and the information feedback process are re-executed until the reservation information matches the reservation standard, and the reservation information is sent to the corresponding service institution.

[0064] It can be understood that when the intention type is a recommendation intention, the historical feature data corresponding to the user is obtained from the user feature database, and the historical feature data records information such as age, gender, and medical history. Then, all business services associated with the target service platform are determined, and the reservation standard corresponding to each business service is identified. At least one recommended business service is determined in all business services, which matches the historical feature data. Then, at least one recommended business service is fed back through voice and / or vision.

[0065] The method provided by the embodiment of the application integrates voiceprint recognition, natural language processing (NLP) and multi-modal interaction technology on the intelligent sound box device, and constructs an interactive service reservation system, which replaces the graphical interface operation of the traditional APP / website and reduces the operation threshold of the user. In the technical application layer, the voiceprint automatic login, information intelligent comparison, multi-round interaction guidance and other processes are automatically executed by the system, which maximally reduces the manual intervention, greatly improves the service reservation efficiency, and effectively avoids the reservation failure problem caused by operation errors. The system finally constructs a closed-loop service process of "voice interaction-intelligent matching-automatic guidance", and realizes the convenience, accuracy and efficiency of service reservation.

[0066] Further, as Figure 1 As a specific implementation of the method, the embodiment of the application provides an interactive service reservation device, such as Figure 9As shown, the device comprises: a collection module 901, a query module 902, a reservation module 903.

[0067] The collection module 901 is configured to collect conversation audio input by a user, determine a user intent corresponding to the conversation audio, and when an intent type of the user intent matches a preset intent type, determine a user voiceprint corresponding to the conversation audio based on voiceprint recognition technology, log into a target service platform according to the user voiceprint, and the target service platform is used for purchasing and reserving a service;

[0068] The query module 902 is configured to, when the intent type is a reservation intent, determine a service to be reserved and reservation information according to the conversation audio, and query a reservation standard corresponding to the service to be reserved, and compare the reservation information with the reservation standard.

[0069] The reservation module 903 is configured to, if the comparison result indicates that the reservation information matches the reservation standard, provide a service institution of the service to be reserved based on a connection of the target service platform, send the reservation information to the service institution, and after receiving reservation success information returned by the service institution, feed back the reservation success information through the smart sound equipment and / or a specified channel.

[0070] In a specific application scenario, the device further comprises a recommendation module.

[0071] The recommendation module is configured to, when the intent type is a recommendation intent, obtain historical feature data corresponding to the user in a user feature database, determine all business services associated with the target service platform, identify a reservation standard corresponding to each business service, determine at least one recommended business service in which a reservation standard matches the historical feature data in all business services, and feed back the at least one recommended business service through voice and / or vision.

[0072] In a specific application scenario, the collection module 901 is configured to collect a user voice signal through a microphone array configured by the smart sound device, obtain the conversation audio, and convert the conversation audio into conversation text through automatic speech recognition (ASR) technology. A natural language processing (NLP) model is used to understand the semantics of the conversation text to determine the user intent. If the intent type of the user intent matches the preset intent type, the voiceprint feature vector of the voice signal is extracted, the voiceprint feature vector is compared with a pre-stored login voiceprint in a voiceprint library, a login voiceprint matching the voiceprint feature vector is determined, and the preset intent type includes the reservation intent and the recommendation intent. Determine the account login information associated with the login voiceprint, and log into the target service platform through the OAuth protocol according to the account login information.

[0073] In a specific application scenario, the query module 902 is configured to perform semantic understanding on the conversation audio by using an NLP model, identify a user intention and entity information, the entity information including the to-be-reserved service and the reservation information, the reservation information including a reservation time, a reservation institution, and a reservation item; if the target service platform provides the to-be-reserved service, query a to-be-reserved item corresponding to the to-be-reserved service, if the to-be-reserved service is associated with a plurality of to-be-reserved items and the reservation information does not include the reservation item, trigger a multi-round interactive clarification process, determine the reservation item, obtain a reservation standard corresponding to the reservation item, and compare the reservation information with the reservation standard; if the to-be-reserved service is associated with a plurality of to-be-reserved items and the reservation information includes the reservation item, obtain a reservation standard corresponding to the reservation item, and compare the reservation information with the reservation standard; if the to-be-reserved service is associated with one to-be-reserved item and the reservation information includes the reservation item, when the to-be-reserved item matches the reservation item, obtain a reservation standard corresponding to the reservation item, and compare the reservation information with the reservation standard; if the target service platform does not provide the to-be-reserved service, generate a reservation failure prompt indicating that the user is not entitled to reserve related services, and feed back the reservation failure prompt through voice and / or vision.

[0074] In a specific application scenario, the query module 902 is further configured to convert the reservation standard into a structured matching rule library by using NLP technology, and standardize the reservation information into structured data including a reservation item, a reservation time, and a reservation institution; obtain historical characteristic data corresponding to the user in a user characteristic database, if the historical characteristic data matches a user attribute recorded in the reservation standard, determine that the user attribute matching is successful, otherwise, determine that the user attribute standard item matching fails; if the reservation time hits a reservable time recorded in the reservation standard, determine that the reservation time matching is successful, otherwise, determine that the reservation time standard item matching fails; if the reservation institution hits a reservable institution recorded in the reservation standard, determine that the reservation institution matching is successful, otherwise, determine that the reservation institution standard item matching fails; query all reservable items associated with the reservation institution in the reservation standard, if the reservation item hits the reservable item, determine that the reservation item matching is successful, otherwise, determine that the reservation item standard item matching fails; if the user attribute, the reservation item, the reservation time, and the reservation institution all match successfully, generate a comparison result indicating that the reservation information matches the reservation standard; if any of the user attribute, the reservation item, the reservation time, and the reservation institution fails to match, generate a comparison result indicating that the reservation information does not match the reservation standard.

[0075] In a specific application scenario, the device further comprises a feedback module.

[0076] The feedback module is configured to, if the comparison result indicates that the reservation information does not match the reservation standard, determine a non-matched reservation standard item, generate a guided interactive session according to the reservation standard item, and output the guided interactive session through voice and / or vision; continuously collect session audio input by a user based on a session context, update target service and reservation information, re-execute the interfacing with the resource institution, the reservation standard comparison and the information feedback process until the reservation information matches the reservation standard, and send the reservation information to the corresponding service institution.

[0077] In a specific application scenario, the device further comprises an interaction module.

[0078] The interaction module is configured to, when continuously collecting session audio input by a user based on a session context, if a semantic ambiguity expression or a multi-modal input instruction conflict is detected, start a multi-round interactive clarification process.

[0079] The device provided by the embodiment of the application integrates voiceprint recognition, natural language processing (NLP) and multi-modal interaction technology on an intelligent sound box device, constructs an interactive service reservation system, replaces the graphical interface operation of a traditional APP / website, and reduces the operation threshold of a user. In terms of technical application, voiceprint automatic login, information intelligent comparison and multi-round interactive guidance are automatically executed by the system, human intervention is minimized, service reservation efficiency is greatly improved, and the problem of reservation failure caused by operation errors is effectively avoided. The system finally constructs a closed-loop service process of "voice interaction-intelligent matching-automatic guidance", and realizes the convenience, accuracy and efficiency of service reservation.

[0080] It should be noted that other corresponding descriptions of each functional unit involved in the interactive service reservation device provided by the embodiment of the application can be referred to the corresponding descriptions in Figure 1 and Figure 8 , which will not be described here.

[0081] To solve the above technical problems, the embodiment of the application further provides a computer device. For details, please refer to Figure 10 , Figure 10 The figure is a basic structure block diagram of the computer device of the embodiment.

[0082] As Figure 10As shown, a schematic diagram of the internal structure of a computer device. The computer device includes a processor, a non-volatile storage medium, a memory and a network interface connected via a system bus. Among them, the non-volatile storage medium of the computer device stores an operating system, a database and computer-readable instructions, and the database may store a control information sequence. When the computer-readable instructions are executed by the processor, the processor can implement a data relationship reconstruction method. The processor of the computer device is used to provide computing and control capabilities to support the operation of the entire computer device. The memory of the computer device may store computer-readable instructions. When the computer-readable instructions are executed by the processor, the processor can execute a data relationship reconstruction method. The network interface of the computer device is used to connect and communicate with the terminal. Those skilled in the art will understand that Figure 10 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0083] In this embodiment, the processor is used to execute Figure 9 The memory stores the program code and various data required to execute the specific functions of the acquisition module 901, query module 902, and reservation module 903. The network interface is used to transmit data between user terminals or servers. The memory in this embodiment stores the program code and data required to execute all submodules in the data relationship reconstruction device. The server can call the server's program code and data to execute the functions of all submodules.

[0084] The present invention also provides a storage medium storing computer-readable instructions. When the computer-readable instructions are executed by one or more processors, the one or more processors execute the steps of the data relationship reconstruction method in any of the above embodiments.

[0085] Those skilled in the art will appreciate that all or part of the processes in the above-described method embodiments can be implemented by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When executed, the program can include the processes in the above-described method embodiments. The aforementioned storage medium can be a non-volatile storage medium such as a magnetic disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

[0086] Those skilled in the art can understand that the steps, measures and schemes in the various operations, methods and processes discussed in the present application can be alternated, changed, combined or deleted. Further, other steps, measures and schemes in the various operations, methods and processes discussed in the present application can also be alternated, changed, rearranged, decomposed, combined or deleted. Further, the steps, measures and schemes in the various operations, methods and processes in the prior art can also be alternated, changed, rearranged, decomposed, combined or deleted.

[0087] The above only describes some embodiments of the present application. It should be pointed out that those skilled in the art can make several improvements and refinements without departing from the principles of the present application, and these improvements and refinements should also be considered as the protection scope of the present application.

Claims

1. An interactive service reservation method, characterized in that: The method is applicable to a smart speaker device, including: Collecting conversation audio input by the user, determining the user intent corresponding to the conversation audio, and if the intent type of the user intent matches a preset intent type, determining the user voiceprint corresponding to the conversation audio based on voiceprint recognition technology, and logging into a target service platform based on the user voiceprint, wherein the target service platform is used to purchase and reserve business services; When the intention type is a reservation intention, determining the service to be reserved and the reservation information based on the conversation audio, querying the reservation standard corresponding to the service to be reserved, and comparing the reservation information with the reservation standard; If the comparison result indicates that the reservation information matches the reservation criteria, the service agency that provides the service to be reserved is connected based on the target service platform, the reservation information is sent to the service agency, and after receiving the reservation success information returned by the service agency, the reservation success information is fed back through the smart audio device and / or designated channel.

2. The method according to claim 1, characterized in that The method further comprises: When the intention type is a recommendation intention, obtaining historical feature data corresponding to the user from a user feature database; Determining all business services associated with the target service platform, identifying the reservation criteria corresponding to each of the business services, and determining at least one business service to be recommended whose reservation criteria matches the historical feature data among all the business services; The at least one business service to be recommended is fed back via voice and / or visual feedback.

3. The method according to claim 1, characterized in that The collecting of the conversation audio input by the user, determining the user intent corresponding to the conversation audio, and when the intent type of the user intent matches a preset intent type, determining the user voiceprint corresponding to the conversation audio based on voiceprint recognition technology, and logging into the target service platform according to the user voiceprint, includes: The microphone array configured in the smart speaker device collects the user's voice signal to obtain the conversation audio, and converts the conversation audio into conversation text through automatic speech recognition technology (ASR) technology; Using an NLP model to perform semantic understanding on the conversation text to determine the user intent and the intent type corresponding to the user intent; If the intention type matches a preset intention type, extracting a voiceprint feature vector of the voice signal, comparing the voiceprint feature vector with a pre-stored login voiceprint in a voiceprint library, and determining a login voiceprint that matches the voiceprint feature vector, wherein the preset intention type includes the reservation intention and the recommendation intention; Determine the account login information associated with the login voiceprint, and log in to the target service platform through the OAuth protocol according to the account login information.

4. The method according to claim 1, wherein The determining of the service to be reserved and the reservation information based on the conversation audio, querying the reservation standard corresponding to the service to be reserved, and comparing the reservation information with the reservation standard includes: Using an NLP model to perform semantic understanding on the conversation audio and identify entity information, the entity information including the service to be reserved and the reservation information, the reservation information including the reservation time, reservation organization, and reservation item; If the target service platform provides the service to be reserved, query the reserved items corresponding to the service to be reserved; if the service to be reserved is associated with multiple reserved items and the reserved item does not exist in the reservation information, trigger multiple rounds of interactive clarification processes to determine the reservation item, obtain the reservation standard corresponding to the reservation item, and compare the reservation information with the reservation standard; if the service to be reserved is associated with multiple reserved items and the reservation item exists in the reservation information, obtain the reservation standard corresponding to the reservation item, and compare the reservation information with the reservation standard; if the service to be reserved is associated with one reserved item and the reservation item exists in the reservation information, then when the reserved item matches the reservation item, obtain the reservation standard corresponding to the reservation item, and compare the reservation information with the reservation standard; If the target service platform does not provide the service to be reserved, a reservation failure prompt is generated to indicate that the user is not authorized to reserve the relevant service, and the reservation failure prompt is fed back via voice and / or visual feedback.

5. The method according to claim 4, characterized in that The comparing the reservation information with the reservation criteria includes: Converting the reservation criteria into a structured matching rule base using NLP technology, and standardizing the reservation information into structured data, including reservation items, reservation times, and reservation agencies; Obtain historical feature data corresponding to the user in a user feature database; if the historical feature data matches the user attributes recorded in the reservation standard, it is determined that the user attributes are matched successfully; otherwise, it is determined that the user attribute standard item matches unsuccessfully; if the reservation time matches the available reservation time recorded in the reservation standard, it is determined that the reservation time is matched successfully; otherwise, it is determined that the reservation time standard item matches unsuccessfully; if the reservation agency matches the available reservation agency recorded in the reservation standard, it is determined that the reservation agency is matched successfully; otherwise, it is determined that the reservation agency standard item matches unsuccessfully; query all available reservation items associated with the reservation agency in the reservation standard; if the reservation item matches the available reservation item, it is determined that the reservation item is matched successfully; otherwise, it is determined that the reservation item standard item matches unsuccessfully; If the user attributes, the reservation item, the reservation time and the reservation organization all match successfully, generating a comparison result indicating that the reservation information matches the reservation criteria; If any of the standard items among the user attributes, the reservation item, the reservation time and the reservation organization fails to match, a comparison result indicating that the reservation information does not match the reservation standard is generated.

6. The method according to claim 1, characterized in that After comparing the reservation information with the reservation criteria, the method further includes: If the comparison result indicates that the reservation information does not match the reservation criteria, determining unmatched reservation criteria items, generating a guided interactive session based on the reservation criteria items, and outputting the guided interactive session via voice and / or visual output; Based on the conversation context, the user's input conversation audio is continuously collected, the target business service and appointment information are updated, and the docking with the resource agency, appointment standard comparison and information feedback process are re-executed until the appointment information matches the appointment standard, and the appointment information is sent to the corresponding service agency.

7. The method according to any one of claims 1 to 6, characterized in that The method further includes: when continuously collecting user input conversation audio based on the conversation context, if semantically ambiguous expressions or multimodal input instruction conflicts are detected, initiating a multi-round interactive clarification process.

8. An interactive service reservation device, characterized in that: include: A collection module is configured to collect conversation audio input by the user, determine the user intent corresponding to the conversation audio, and, when the intent type of the user intent matches a preset intent type, determine the user voiceprint corresponding to the conversation audio based on voiceprint recognition technology, and log in to a target service platform based on the user voiceprint, wherein the target service platform is used to purchase and reserve business services; A query module is used to determine the service to be reserved and the reservation information based on the conversation audio when the intention type is a reservation intention, and to query the reservation standard corresponding to the service to be reserved, and compare the reservation information with the reservation standard; The reservation module is used to connect with the service agency that provides the service to be reserved based on the target service platform if the comparison result indicates that the reservation information matches the reservation standard, send the reservation information to the service agency, and after receiving the reservation success information returned by the service agency, feedback the reservation success information through the smart audio device and / or designated channel.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 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 7 are implemented.

Images