CN116774818A

CN116774818A - Virtual attendant interaction method, electronic device, storage medium and program product

Info

Publication number: CN116774818A
Application number: CN202310573523.4A
Authority: CN
Inventors: 黄晨升; 谭佩桢; 张晨露; 柴金祥
Original assignee: Shanghai Movu Technology Co Ltd; Mofa Shanghai Information Technology Co Ltd
Current assignee: Shanghai Movu Technology Co Ltd; Mofa Shanghai Information Technology Co Ltd
Priority date: 2023-05-21
Filing date: 2023-05-21
Publication date: 2023-09-19

Abstract

The application provides a virtual attendant interaction method, electronic equipment, a computer readable storage medium and a computer program product, wherein the method comprises the following steps: receiving an access request from first terminal equipment, and establishing communication connection between the first terminal equipment and a target server; receiving input information of a user by using the first terminal equipment; determining one of a plurality of service items as a target service item of the user; acquiring a first interactive video corresponding to the target service item from a question-answer library corresponding to the virtual service personnel; and playing the first interactive video by using the first terminal equipment, wherein in the first interactive video, the virtual service personnel introduces a business process and/or a business guide corresponding to the target service item. The application provides the interactive video related to the target service matters, so that a user can know the required service contents and operation steps more clearly, and the user satisfaction and the business efficiency are improved.

Description

Virtual attendant interaction method, electronic device, storage medium and program product

Technical Field

The present application relates to the field of virtual man and artificial intelligence, and more particularly, to a virtual attendant interaction method, an electronic device, a computer readable storage medium, and a computer program product.

Background

With the rapid development of internet technology and the acceleration of people's life pace, more and more service matters start to shift online. To improve user convenience and business efficiency, enterprises or government agencies have begun to use virtual service personnel to provide online consultation and business guides. The virtual service personnel can provide real-time interaction according to the requirements of users, so that the questions of the users can be effectively answered, and the satisfaction degree of the users is improved.

The existing virtual attendant interaction method still has the defects in the aspects of processing user input information and providing personalized services. Based on this, the present application provides a virtual attendant interaction method, an electronic device, a computer readable storage medium and a computer program product, to improve the prior art.

Disclosure of Invention

The application aims to provide a virtual service personnel interaction method, electronic equipment, a computer readable storage medium and a computer program product, which can quickly determine target service matters according to input information and provide an interaction video related to the target service matters so that a user can know required service contents and operation steps more clearly, and further, the user satisfaction and the business handling efficiency are improved.

The application adopts the following technical scheme:

in a first aspect, the present application provides a virtual attendant interaction method, the method comprising:

receiving an access request from first terminal equipment, and establishing communication connection between the first terminal equipment and a target server, wherein the target server is used for providing a virtual service personnel interaction function;

receiving input information of a user by using the first terminal device, wherein each input information is text information, voice information or image information;

determining one service item from a plurality of service items as a target service item of the user according to the input information;

acquiring a first interactive video corresponding to the target service item from a question-answer library corresponding to the virtual service person, wherein the question-answer library comprises first interactive videos corresponding to a plurality of service items;

and playing the first interactive video by using the first terminal equipment, wherein in the first interactive video, the virtual service personnel introduces a business process and/or a business guide corresponding to the target service item.

The beneficial effect of this technical scheme lies in: by adopting a mode of establishing communication connection between the first terminal equipment and the target server, the interaction function of virtual service personnel is realized, the instantaneity and the stability of the system are improved, and smoother online consultation experience is provided for users. Through receiving user's input information, including text information, voice information and image information, can satisfy different user demands, provide more convenient interactive mode, reinforcing user experience. By determining the target service item from the plurality of service items according to the information input by the user, the user demand can be rapidly identified, and the pertinence and the accuracy of the virtual service personnel are improved. The first interactive video corresponding to the target service item is obtained from the question-answer library corresponding to the virtual service personnel, and detailed interactive video related to the target service item can be provided, so that a user can more intuitively know related information of the target service item, and user satisfaction and working efficiency are improved. In the first interactive video, the virtual service personnel introduces the corresponding business flow and/or business guide of the target service item, detailed guiding information is provided for the user, the user can know the business flow and the requirements more clearly, and the business efficiency and the business accuracy are improved.

In summary, the virtual service personnel interaction method provided by the application can receive various types of input information, quickly determine target service matters, provide detailed business processes and business guides, and effectively improve convenience, business efficiency and satisfaction of users. Meanwhile, the method ensures real-time performance and stability in the interaction process of the user and the virtual service personnel, ensures that the user can find out required business processes and business guides in a short time, and can better adapt to the requirements of the user, improve the dependence and satisfaction of the user on online service and improve the user experience compared with the prior art. Therefore, the technical scheme has higher application value and practicability in the field of virtual service personnel interaction, and provides an effective way for improving customer service and user experience for enterprises and government departments.

In some possible implementations, the determining, according to the input information, one service item from a plurality of service items as a target service item of the user includes:

extracting semantic information from the input information by using a semantic extraction model corresponding to the input information;

Obtaining the similarity between the semantic information and each service item, and taking the maximum value of the similarity as the highest similarity;

and when the highest similarity is larger than a preset similarity, taking the service item corresponding to the highest similarity as the target service item.

The beneficial effect of this technical scheme lies in: by extracting semantic information from the input information by using the semantic extraction model, the user requirement can be more accurately understood, and the response accuracy of the virtual service personnel is effectively improved. By calculating the similarity between the semantic information and each service item, the service item which is most matched with the user input information can be quickly found, and the pertinence and the accuracy of the virtual service personnel are further improved. By setting the preset similarity, misjudgment and mismatching can be effectively avoided, and only highly matched service matters are ensured to be provided for the user as target service matters. The technical scheme enhances the accuracy and reliability of the virtual service personnel in processing complex problems and fuzzy input, thereby improving the user satisfaction and the work efficiency. In summary, by adopting the semantic extraction model, calculating the similarity and setting the preset similarity, the technical scheme can more accurately determine the target service items of the user and provide more accurate and personalized service for the user. The implementation method has higher practicability and efficiency in the field of virtual service personnel interaction, is beneficial to further improving user experience and satisfaction, and enables enterprises and government departments to better meet the demands of various groups of people.

In some possible implementations, the determining, according to the input information, one service item from a plurality of service items as a target service item of the user further includes:

when the highest similarity is not greater than the preset similarity, playing a second interactive video by using the first terminal equipment so as to guide the user to click a form button and fill in a form;

when a clicking operation for the form button is received, displaying the form by using the first terminal equipment, wherein the form is provided with a service item input box and/or a service item drop-down menu;

and acquiring the target service items of the user according to the filling result of the form.

The beneficial effect of this technical scheme lies in: when the highest similarity is not greater than the preset similarity, the second interactive video is played to guide the user to click the form button and fill in the form, so that more information about the user requirement can be collected, the detailed requirement of the user can be obtained more effectively, the accuracy of the target service item can be improved, and therefore more accurate and personalized service can be provided for the user. The form is provided with a service item input box and/or a service item drop-down menu, so that a user has greater flexibility in providing requirements, and is convenient for the user to select and input the required service items. Under the condition that the virtual service personnel cannot accurately identify the user demands, an effective supplementary means is provided, and the user satisfaction and the office efficiency are improved. In summary, according to the technical scheme, when the highest similarity does not meet the preset condition, the user is guided to fill in the form to acquire more information about the user requirement, so that the actual requirement of the user is better met, and the overall working efficiency and the user satisfaction are improved.

In some possible implementations, the process of obtaining the similarity of the semantic information to each of the service items includes:

for each of the service matters, the following processing is performed:

inputting the semantic information and a preset text corresponding to the service item into a similarity model to obtain the similarity of the semantic information and the preset text, wherein the similarity is used as the similarity of the semantic information and the service item;

the training process of the similarity model comprises the following steps:

acquiring a training set, wherein the training set comprises a plurality of training data, and each training data comprises sample semantic information, a sample text and annotation data of similarity of the sample semantic information and the sample text;

for each of the training data, the following processing is performed:

inputting sample semantic information and sample text in the training data into a preset deep learning model to obtain prediction data of similarity of the sample semantic information and the sample text;

updating model parameters of the deep learning model according to the sample semantic information, the prediction data and the labeling data of the similarity of the sample text;

Detecting whether a preset training ending condition is met; if yes, taking the trained deep learning model as the similarity model; if not, continuing to train the deep learning model by using the next training data.

The beneficial effect of this technical scheme lies in: for each service item, the semantic information and the preset text are input into a similarity model, so that the similarity between the semantic information and each service item can be calculated more accurately, and the matching accuracy of virtual service personnel is improved. The training process of the similarity model adopts a deep learning model, and training is carried out by utilizing sample semantic information and sample text in training data, so that the similarity model can better capture the association between the semantic information and the text, and the accuracy of similarity calculation is improved. By detecting the preset training ending condition, the similarity model can be trained on the premise of meeting a certain precision requirement, so that the generalization capability and practicability of the model are improved. According to the technical scheme, the similarity calculation is performed by using the deep learning model, so that service matters in different fields can be better adapted, and the universality and the application value of virtual service personnel in various scenes are improved. The method has higher accuracy and reliability when processing the input information of the user, thereby improving the satisfaction degree and the working efficiency of the user and enabling enterprises and government departments to better meet the demands of clients. In summary, according to the technical scheme, the similarity model is trained through the deep learning model, so that more accurate similarity calculation between semantic information and service matters is realized, and further accuracy and practicability of virtual service personnel in determining target service matters are improved.

In some possible implementations, the method further includes:

receiving feedback information of the user by using the first terminal equipment, wherein each feedback information is text information, voice information or image information;

identifying emotion categories of the user according to the feedback information;

detecting whether the user is satisfied with the service according to the emotion type of the user;

if not, adding one to the accumulated times;

when the accumulated times are larger than preset times, carrying out zero clearing processing on the accumulated times, and sending prompt information to second terminal equipment of preset personnel corresponding to the target service items so as to prompt the preset personnel to update the similarity model or update preset texts corresponding to the target service items.

The beneficial effect of this technical scheme lies in: by receiving feedback information of the user and identifying emotion categories of the user, satisfaction degree of the user on services provided by virtual service personnel can be better known, and further optimization of service quality is facilitated. According to the emotion type of the user, whether the user is satisfied with the service or not is detected, and service matters with problems can be found in time, so that the method is improved in a targeted manner. When the accumulated times are greater than the preset times, prompt information is sent to a second terminal device of the preset personnel, so that the related personnel can be reminded of updating the similarity model or adjusting the preset text, and the matching accuracy and the user satisfaction of the virtual service personnel are improved. The method strengthens the accuracy and reliability of the virtual service personnel when processing the user demands, thereby improving the user satisfaction degree and the office efficiency and enabling enterprises and government departments to better meet the client demands. Through user feedback and emotion recognition, the technical scheme is beneficial to collecting precious user opinions and suggestions, provides powerful support for further optimizing the working effect of virtual service personnel, ensures that related personnel pay attention to the problem of user satisfaction in time, and prevents the quality of service from sliding down. In conclusion, the technical scheme can continuously optimize the service quality and the matching accuracy of the virtual service personnel by collecting the user feedback information, identifying the emotion type of the user and sending the prompt information.

In some possible implementations, the method further includes:

displaying a plurality of second interactive videos related to the first interactive video by using the first terminal equipment, wherein each second interactive video is used for introducing notes, common problems or related service matters of the target service matters;

and when a clicking operation for one of the second interactive videos is received, playing the clicked second interactive video by using the first terminal equipment.

The beneficial effect of this technical scheme lies in: by displaying a plurality of second interactive videos related to the first interactive video on the first terminal device, richer and more comprehensive service information can be provided for the user, so that the user can be helped to better know the notice, common problems or related service matters of the target service matters. The user can select to watch the second interactive video related to the target service item according to the own requirements and interests, so that the initiative and the interactivity of the user in the process of acquiring the service information are improved, and the user satisfaction is improved. The method provides a more convenient and personalized information acquisition mode for the user, and is beneficial to meeting the requirements and preferences of different users.

In some possible implementations, in the first interactive video, the virtual attendant directs the user to click a document download button to download a certification document;

the method further comprises the steps of:

and when a clicking operation for the document downloading button is received, downloading the proving document corresponding to the document downloading button to the first terminal equipment.

The beneficial effect of this technical scheme lies in: in the first interactive video, the virtual service personnel guides the user to click a document downloading button, so that the user can easily acquire the required proving document, and the user experience and satisfaction are improved. By responding to the clicking operation of the user, the document is downloaded to the first terminal device, so that convenience and high efficiency of the document downloading function are realized, and time and energy of the user are saved. The technical scheme simplifies the acquisition process of the proving document, thereby improving the working efficiency. In the interaction process of the virtual service personnel, the downloading function of the proof document is realized, the service content provided by the virtual service personnel is further enriched, and the user can complete more service matters on one platform.

In some possible implementations, when receiving a click operation for the document download button, downloading a proof document corresponding to the document download button to the first terminal device, including:

When a clicking operation for the document downloading button is received, displaying a document downloading interface corresponding to the document downloading button by using the first terminal equipment, wherein the document downloading interface is provided with an identity recognition button;

when a clicking operation aiming at the identity recognition button is received, acquiring identity information to be recognized of the user by using the first terminal equipment;

detecting whether the identity information to be identified is matched with authenticated identity information of the user;

and if the two types of the data are matched, downloading the certification document to the first terminal equipment.

The beneficial effect of this technical scheme lies in: after clicking operation for the document downloading button is received, displaying a document downloading interface, wherein the document downloading interface is provided with an identity recognition button for triggering an identity authentication step for a user, so that a safe identity authentication mechanism is provided, and downloading safety and privacy protection of a proving document are ensured. After clicking the identity recognition button, the first terminal device acquires the identity information to be recognized of the user, and is helpful to ensure that the certification document acquired by the user accords with the identity of the user. By detecting the matching condition of the identity information to be identified and the authenticated identity information, the malicious user can be prevented from downloading the proving document of other people to a certain extent, and the information security is improved. If the identity information to be identified is matched with the authenticated identity information, the proving document is downloaded to the first terminal equipment, so that a user can successfully acquire the required proving document, and the user experience and satisfaction are improved.

In some possible implementations, said downloading the certification document to the first terminal device if there is a match, includes:

if the identification information is matched with the identification information, inquiring a user information database corresponding to the target service item according to the authenticated identification information, so as to generate the proving document according to the inquiring result;

and downloading the proving document to the first terminal equipment.

The beneficial effect of this technical scheme lies in: and inquiring a user information database corresponding to the target service item according to the authenticated identity information, generating a proving document according to the inquiring result, ensuring that the generated proving document is based on the real information of the user, and ensuring timeliness, reliability and accuracy of the generated proving document. In the interaction process of the virtual service personnel, the generation and downloading functions of the certification document are realized, and the service content provided by the virtual service personnel is further enriched.

In some possible implementations, the method further includes:

and acquiring one or more recommended policies corresponding to the user according to the authenticated identity information of the user and the interaction data between the user and the virtual service personnel, and displaying the recommended policies on the first terminal equipment in a list form.

The beneficial effect of this technical scheme lies in: by combining authenticated identity information of the user and interaction data between the user and the virtual service personnel, one or more recommendation policies corresponding to the user can be more accurately obtained, so that pertinence and accuracy of recommendation are improved. The recommended policies are displayed on the first terminal device in a list form, so that a clear and visual information display mode is provided for the user, and the user can quickly browse and know the policies related to the user. The method realizes the policy recommendation function, can help the user to know the policy information related to the user in time, improves the attention degree and the cognition degree of the user on the policy, is beneficial to the benefit of the user from the policy, and avoids the user from missing the policy benefit related to the user. By recommending the policies related to the identity and the requirements of the user, the policy coverage and implementation effect can be improved, and the overall benefit of policy implementation can be improved.

In a second aspect, the present application provides an electronic device comprising a memory storing a computer program and at least one processor configured to implement the following steps when executing the computer program:

In a third aspect, the present application provides a computer-readable storage medium storing a computer program which, when executed by at least one processor, performs the steps of any of the methods or performs the functions of any of the electronic devices described above.

In a fourth aspect, the application provides a computer program product comprising a computer program which, when executed by at least one processor, performs the steps of any of the methods or performs the functions of any of the electronic devices described above.

Drawings

The application is further described below with reference to the drawings and the detailed description.

Fig. 1 is a flow chart of a virtual attendant interaction method according to an embodiment of the present application.

Fig. 2 is a schematic flow chart of acquiring a target service item according to an embodiment of the present application.

Fig. 3 is a schematic flow chart of prompting a preset person to update according to an embodiment of the present application.

Fig. 4 is a schematic flow chart of a downloading certification document according to an embodiment of the present application.

Fig. 5 is a block diagram of an electronic device according to an embodiment of the present application.

Fig. 6 is a schematic structural diagram of a computer program product according to an embodiment of the present application.

Detailed Description

The technical scheme of the present application will be described below with reference to the drawings and the specific embodiments of the present application, and it should be noted that, on the premise of no conflict, new embodiments may be formed by any combination of the embodiments or technical features described below.

In embodiments of the application, words such as "exemplary" or "such as" are used to mean serving as an example, instance, or illustration. Any implementation or design described as "exemplary" or "e.g." in the examples of this application should not be construed as preferred or advantageous over other implementations or designs. Rather, the use of words such as "exemplary" or "such as" is intended to present related concepts in a concrete fashion.

The first, second, etc. descriptions in the embodiments of the present application are only used for illustration and distinction of description objects, and no order division is used, nor does it represent a particular limitation on the number in the embodiments of the present application, nor should it constitute any limitation on the embodiments of the present application.

The technical field and related terms of the embodiments of the present application are briefly described below.

The virtual objects include virtual humans, virtual animals, virtual cartoon figures, and the like. The virtual person is a personified image constructed by CG technology and operated in a code form, and has various interaction modes such as language communication, expression, action display and the like. The technology of virtual persons has been rapidly developed in the field of artificial intelligence and has been applied in many technical fields such as video, media, games, finance, travel, education, medical treatment, etc., and not only can a virtual host, a virtual anchor, a virtual idol, a virtual customer service, a virtual lawyer, a virtual serviceman, a virtual doctor, a virtual instructor, a virtual assistant, etc. be customized, but also a video can be generated through text or audio one-key. In the virtual people, the service type virtual people mainly have the functions of replacing real people to serve and provide daily accompaniment, are the virtualization of service type roles in reality, and have the industrial value of mainly reducing the cost of the existing service type industry and enhancing the cost reduction of the stock market.

Artificial intelligence (Artificial Intelligence, AI) is the theory, method, technique and application system that uses a digital computer or a machine controlled by a digital computer to simulate, extend and extend human intelligence, sense the environment, acquire knowledge and use the knowledge to obtain optimal results. In other words, artificial intelligence is an integrated technology of computer science that attempts to understand the essence of intelligence and to produce a new intelligent machine that can react in a similar way to human intelligence. The design principle and the implementation method of various intelligent machines are researched by artificial intelligence, so that the machines have the functions of perception, reasoning and decision. The artificial intelligence technology is a comprehensive subject, and relates to the technology with wide fields, namely the technology with a hardware level and the technology with a software level. Artificial intelligence infrastructure technologies generally include technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing technologies, operation/interaction systems, mechatronics, and the like. The artificial intelligence software technology mainly comprises a computer vision technology, a voice processing technology, a natural language processing technology, machine learning/deep learning, automatic driving, intelligent traffic and other directions.

Machine Learning (ML) is a multi-domain interdisciplinary, involving multiple disciplines such as probability theory, statistics, approximation theory, convex analysis, algorithm complexity theory, etc. The computer program may learn experience E given a certain class of tasks T and performance metrics P, and increase with experience E if its performance in task T happens to be measured by P. Machine learning is specialized in studying how a computer simulates or implements learning behavior of a human to acquire new knowledge or skills, reorganizing existing knowledge structures to continually improve its own performance. Machine learning is the core of artificial intelligence, a fundamental approach to letting computers have intelligence, which is applied throughout various areas of artificial intelligence.

Deep learning is a special machine learning by which the world is represented using a hierarchy of nested concepts, each defined as being associated with a simple concept, and achieving great functionality and flexibility, while a more abstract representation is computed in a less abstract way. Machine learning and deep learning typically include techniques such as artificial neural networks, belief networks, reinforcement learning, transfer learning, induction learning, teaching learning, and the like.

The virtual object interaction application is used for providing virtual object interaction functions. Virtual objects can simulate human communication and behavior and interact with users. Such software (referred to as virtual human interactive applications) is typically driven by artificial intelligence and natural language processing techniques and is capable of interacting with a user by means of text, speech, images, forms, etc.

The existing virtual attendant interaction method still has the defects in the aspects of processing user input information and providing personalized services. For example, some existing virtual attendant interaction methods can only process text information, and cannot recognize and process voice or image information. In addition, some existing virtual service personnel lack pertinence in providing services, and the target service items cannot be determined quickly according to the information input by the user, so that the user needs to spend more time in interaction with the virtual service personnel, and the working efficiency is reduced. Meanwhile, the existing virtual service personnel interaction method cannot provide clear and visual working procedures and working guides when the interaction video is played, and user satisfaction is further reduced.

Based on this, the present application provides a virtual attendant interaction method, an electronic device, a computer readable storage medium and a computer program product, to improve the prior art.

The scheme provided by the embodiment of the application relates to technologies such as virtual man, interactive design, artificial intelligence, 3D modeling, cloud computing and the like, and is specifically described by the following embodiment. The following description of the embodiments is not intended to limit the preferred embodiments.

(virtual attendant interaction method)

Referring to fig. 1, fig. 1 is a flow chart of a virtual attendant interaction method according to an embodiment of the present application.

The embodiment of the application provides a virtual attendant interaction method, which comprises the following steps:

step S101: receiving an access request from first terminal equipment, and establishing communication connection between the first terminal equipment and a target server, wherein the target server is used for providing a virtual service personnel interaction function;

step S102: receiving input information of a user by using the first terminal device, wherein each input information is text information, voice information or image information;

step S103: determining one service item from a plurality of service items as a target service item of the user according to the input information;

Step S104: acquiring a first interactive video corresponding to the target service item from a question-answer library corresponding to the virtual service person, wherein the question-answer library comprises first interactive videos corresponding to a plurality of service items;

step S105: and playing the first interactive video by using the first terminal equipment, wherein in the first interactive video, the virtual service personnel introduces a business process and/or a business guide corresponding to the target service item.

The virtual attendant interaction method can be operated on the electronic equipment, the electronic equipment and the first terminal equipment (used by a user) can be independent, and the electronic equipment and the first terminal equipment can be integrated. When the electronic device and the first terminal device are independent, the electronic device may be a computer, a server (including a cloud server), or the like having computing power. The first terminal device is not limited in the embodiment of the present application, and may be, for example, an intelligent first terminal device having a display screen and a speaker, such as a mobile phone, a tablet computer, a notebook computer, a desktop computer, an intelligent wearable device, or the first terminal device may be a workstation or a console having a display screen and a speaker. The display screen may be a touch display screen or a non-touch display screen.

In an embodiment of the present application, the virtual attendant comprises one or more of a virtual person, a virtual animal, and a virtual cartoon character. As one example, the virtual attendant is a virtual person "JING" (chinese name: mirror). As one example, the name of the virtual attendant interactive application is "virtual attendant, let you do no more difficult.

The embodiment of the application does not limit the plurality of service matters, and the plurality of service matters of the enterprise can comprise financial services, maintenance, goods return, complaint treatment, safety production management, emergency treatment of safety accidents, recruitment, training, performance assessment and the like. The plurality of services of the government agency may include, for example, enterprise registration, land use applications, security production approval, medical insurance reimbursement, social insurance card reimbursement, tax reduction and exemption applications, talent reimbursement applications, and the like.

The virtual service personnel may provide the user with a variety of business processes and business guides for the services, e.g., the virtual service personnel of the enterprise may provide product or service application business processes and business guides, after-sales service business processes and business guides, complaint handling business processes and business guides, security production business processes and business guides, human resource management business processes and business guides, etc. The product or service application business processes and business guides are mainly aimed at products or services provided by enterprises, such as insurance products, financial services and the like. The enterprise may formulate related processes and guidelines to help customers understand the application processes, conditions, and criteria of the product or service, etc. The after-sales service business process business guide mainly aims at after-sales services provided by enterprises, such as maintenance, goods return and replacement, and the like. The enterprise may formulate related processes and guidelines to assist the customer in understanding the flow, conditions, criteria, etc. of the after-market service application. The complaint handling class business flow business guide mainly aims at the complaint handling of clients, such as complaint handling, complaint investigation, complaint handling and the like. Enterprises can formulate relevant flows and guidelines to help customers understand the flows and relevant regulations of complaint treatments. The safety production process handling guideline mainly aims at safety production management, safety accident emergency treatment and the like of enterprises. The enterprise can formulate related processes and guidelines to help staff and customers to understand the enterprise's safety production management system and emergency treatment measures. The manual resource management process business guide mainly aims at the manual resource management of enterprises, such as recruitment, training, performance assessment and the like. The enterprise will formulate relevant processes and guidelines to help the staff to understand the enterprise's human resource management system and relevant policy specifications.

Virtual service personnel of government departments can provide administrative approval business processes and business guides, public service business processes and business guides, policy and regulation business processes and business guides, government public business processes and business guides and the like. The administrative approval process and the administrative approval manual are mainly aimed at matters needing administrative approval, such as enterprise registration, land use application, safe production approval and the like. The government can formulate related processes and guidelines to help the applicant understand the application processes and reduce problems occurring in the application process. Public service class business processes and business guides are mainly directed to some basic life services of the public, such as medical services, educational services, social security services, and the like. The government may make related processes and guidelines to assist the citizen in knowing the application processes, conditions, criteria, etc. of the related services. The policy and regulation class business processes and business guides are mainly directed to some policy and regulations of government attendance, such as tax offers, subsidy policies, etc. Governments may formulate related procedures and guidelines, assist related units and individuals in understanding the application conditions, application procedures, etc. of the policy regulations. The government open class business processes and business guides are mainly a form of government open government information, such as the organization, responsibility, workflow, etc. of open government departments. The government may develop related processes and guidelines to assist the citizens in understanding the internal operations of government agencies and the implementation of related policies.

The target server may run one or more applications for providing interactive functionality for virtual attendant. These applications may be written using one or more programming languages, such as Java, python, node. Js, etc., and may utilize various frameworks and libraries to implement various functions, such as natural language processing, speech-to-text, image recognition, etc. To improve availability and performance of the target server, a load balancer may be used to distribute requests across multiple servers, and a failover mechanism may be used to automatically switch to a standby server in the event of a server failure. The electronic device and the target server can be independent, and the electronic device and the target server can be integrated.

The question and answer library corresponding to the virtual service personnel needs to be stored on one or more servers, and the servers can store videos in a cloud storage mode or an own storage mode.

The question-answering library corresponding to the virtual service personnel stores first interactive videos corresponding to a plurality of service matters, for example, the question-answering library corresponding to the virtual service personnel of a government department can store first interactive videos corresponding to service matters such as enterprise registration, land use application, safe production approval, medical insurance reimbursement, social insurance card reimbursement, tax deduction application, talent reimbursement application and the like.

Therefore, by adopting a mode of establishing communication connection between the first terminal equipment and the target server, the interaction function of the virtual service personnel is realized, the instantaneity and the stability of the system are improved, and smoother online consultation experience is provided for the user. Through receiving user's input information, including text information, voice information and image information, can satisfy different user demands, provide more convenient interactive mode, reinforcing user experience. By determining the target service item from the plurality of service items according to the information input by the user, the user demand can be rapidly identified, and the pertinence and the accuracy of the virtual service personnel are improved. The first interactive video corresponding to the target service item is obtained from the question-answer library corresponding to the virtual service personnel, and detailed interactive video related to the target service item can be provided, so that a user can more intuitively know related information of the target service item, and user satisfaction and working efficiency are improved. In the first interactive video, the virtual service personnel introduces the corresponding business flow and/or business guide of the target service item, detailed guiding information is provided for the user, the user can know the business flow and the requirements more clearly, and the business efficiency and the business accuracy are improved.

Illustrating the user interaction with the virtual attendant "small a".

A user accesses a virtual serviceman "small a" of a government department (i.e., accesses a target server for providing a "small a" interactive function) through a browser or a dedicated government service application (e.g., a "government service") using a smart phone or computer (i.e., a first terminal device).

The user may interact with the virtual attendant "small a" through text, voice, or images, such as the user entering text information: "I want to know the flow of enterprise registration". That is, in this example, the user's target service item is enterprise registration.

And searching a first interactive video related to enterprise registration in a question-answer library corresponding to the small A of the virtual service personnel according to input information provided by the user, and playing the first interactive video.

In the first interactive video played, the virtual attendant "small a" details the business process of enterprise registration, for example: a. preparing related materials, such as business license copies and the like; b. logging in a government website to carry out enterprise name pre-approval; c. submitting an enterprise registration application form and related materials; d. paying a registered capital; e. get business licenses, etc.

In addition to the business process, the virtual attendant "small A" also provides business guides for business registration in the first interactive video, such as: a. the conditions which the enterprise registration needs to meet; b. relevant laws and policies of enterprise registration; c. a bill of materials to be submitted; d. time requirements and costs for enterprise registration; e. common questions and solutions, etc.

Through the steps, the virtual service personnel 'small A' realizes effective interaction with the user, and helps the user to know the relevant flow and guide of enterprise registration. The user can more easily complete the enterprise registration matters according to the provided information.

Referring to fig. 2, fig. 2 is a schematic flow chart of acquiring a target service item according to an embodiment of the present application.

In some embodiments, the determining, according to the input information, one service item from a plurality of service items as a target service item of the user (i.e. step S103) includes:

In some embodiments, when the input information is text information, the semantic extraction model corresponding to the input information is a pre-trained language model based on deep learning; when the input information is voice information, the semantic extraction model corresponding to the input information comprises a voice-to-text model based on deep learning and a pre-training language model based on deep learning; when the input information is image information, the semantic extraction model corresponding to the input information is a semantic segmentation model based on deep learning.

Thus, a deep learning-based model is used to extract semantic features of the input information. For different types of input information, different deep learning models are adopted for semantic extraction, including a pre-training language model, a voice-to-text model, a semantic segmentation model and the like.

For text information, pre-trained language models, such as BERT, GPT, etc., are employed to extract semantic features of the input text. The models can learn the structure and semantic information of the language by pre-training a large amount of texts, so that the semantic information of the input texts can be effectively extracted.

For voice information, a voice-to-text model based on deep learning, such as CTC, transformer, is adopted to convert the voice information into text information, and a pre-training language model is adopted to extract semantic features of the text information. Thus, semantic information related to the input information can be extracted from the voice information.

For image information, a semantic segmentation model based on deep learning, such as UNet, deep Lab and the like, is adopted to segment the image, and semantic information corresponding to each pixel point is extracted. Thus, semantic information related to the input information can be accurately extracted from the image.

The method has the advantages that semantic features of input information can be extracted more accurately by using a semantic extraction model based on deep learning, so that accuracy and efficiency of semantic understanding are improved; the pre-training language model can improve the natural language processing capacity, including text classification, emotion analysis, machine translation and other aspects; by adopting different deep learning models, the multi-mode information can be processed, including text information, voice information, image information and the like, so that more comprehensive information processing is realized.

The preset similarity refers to a preset similarity threshold value to judge the matching degree of the input information and the target service item. The preset similarity is not limited in the embodiment of the present application, and may be, for example, 0.5, 0.65, 0.75, 0.8, 0.95, etc.

Therefore, the semantic information is extracted from the input information by using the semantic extraction model, so that the requirements of the user can be more accurately understood, and the response accuracy of the virtual service personnel is effectively improved. By calculating the similarity between the semantic information and each service item, the service item which is most matched with the user input information can be quickly found, and the pertinence and the accuracy of the virtual service personnel are further improved. By setting the preset similarity, misjudgment and mismatching can be effectively avoided, and only highly matched service matters are ensured to be provided for the user as target service matters. The technical scheme enhances the accuracy and reliability of the virtual service personnel in processing complex problems and fuzzy input, thereby improving the user satisfaction and the work efficiency. In summary, by adopting the semantic extraction model, calculating the similarity and setting the preset similarity, the technical scheme can more accurately determine the target service items of the user and provide more accurate and personalized service for the user. The implementation method has higher practicability and efficiency in the field of virtual service personnel interaction, is beneficial to further improving user experience and satisfaction, and enables enterprises and government departments to better meet the demands of various groups of people.

With continued reference to fig. 2, in some embodiments, the determining, according to the input information, one service item from a plurality of service items as a target service item of the user (i.e. step S103) further includes:

Therefore, when the highest similarity is not greater than the preset similarity, the second interactive video is played to guide the user to click the form button and fill in the form, so that more information about the user requirement can be collected, the detailed requirement of the user can be acquired more effectively, the accuracy of the target service item can be improved, and more accurate and personalized service can be provided for the user. The form is provided with a service item input box and/or a service item drop-down menu, so that a user has greater flexibility in providing requirements, and is convenient for the user to select and input the required service items. Under the condition that the virtual service personnel cannot accurately identify the user demands, an effective supplementary means is provided, and the user satisfaction and the office efficiency are improved. In summary, according to the technical scheme, when the highest similarity does not meet the preset condition, the user is guided to fill in the form to acquire more information about the user requirement, so that the actual requirement of the user is better met, and the overall working efficiency and the user satisfaction are improved.

Illustrating the user interaction with the virtual attendant "small a".

The user accesses the government virtual attendant "small a" by a browser or a dedicated government service application (e.g., a "government service") using a smart phone or computer (i.e., a first terminal device).

The user may interact with the virtual attendant "small a" by voice, e.g., the user enters a piece of voice, the content being: "I want to know the information set up by the business. "

Semantic information is extracted from input information by using a semantic extraction model corresponding to the semantic information, such as: "company setup", "information".

The virtual attendant "small a" may provide a number of service items, such as enterprise registration, business changes, tax registration, etc.

The similarity between the semantic information extracted from the input information and each service item is calculated, for example: "Enterprise registration: 0.90, business change: 0.65 tax registration: 0.70". In this example, the highest similarity is with the enterprise registration service transaction, i.e., 0.90. The preset similarity is, for example, 0.75.

When the highest similarity is greater than the preset similarity (in this example, 0.90> 0.75), the service item corresponding to the highest similarity is used as the target service item, i.e. the enterprise registration.

If the highest similarity is not greater than the preset similarity, the first terminal device plays a second interactive video of the small A of the virtual service personnel, and guides the user to click a form button and fill in the form. After the user clicks the form button, the first terminal device displays a form with a service item input box and/or a service item drop-down menu to provide the user with more detailed demand information. The user fills in corresponding information according to the form, for example: "Enterprise registration". And determining that the target service item of the user is enterprise registration according to the filling result.

Through the steps, when the user provides fuzzy information, the virtual service personnel 'small A' can finally determine the target service items of the user by calculating the similarity and guiding the user to fill in the form. This helps to provide more accurate services, meeting user needs.

In some embodiments, the process of obtaining the similarity of the semantic information to each of the service matters comprises:

for each of the service matters, the following processing is performed:

The training process of the similarity model comprises the following steps:

for each of the training data, the following processing is performed:

Therefore, for each service item, the semantic information and the preset text are input into the similarity model, so that the similarity between the semantic information and each service item can be calculated more accurately, and the matching accuracy of the virtual service personnel is improved. The training process of the similarity model adopts a deep learning model, and training is carried out by utilizing sample semantic information and sample text in training data, so that the similarity model can better capture the association between the semantic information and the text, and the accuracy of similarity calculation is improved. By detecting the preset training ending condition, the similarity model can be trained on the premise of meeting a certain precision requirement, so that the generalization capability and practicability of the model are improved. According to the technical scheme, the similarity calculation is performed by using the deep learning model, so that service matters in different fields can be better adapted, and the universality and the application value of virtual service personnel in various scenes are improved. The method has higher accuracy and reliability when processing the input information of the user, thereby improving the satisfaction degree and the working efficiency of the user and enabling enterprises and government departments to better meet the demands of clients. In summary, according to the technical scheme, the similarity model is trained through the deep learning model, so that more accurate similarity calculation between semantic information and service matters is realized, and further accuracy and practicability of virtual service personnel in determining target service matters are improved.

Referring to fig. 3, fig. 3 is a schematic flow chart for prompting a preset person to update according to an embodiment of the present application.

In some embodiments, the method further comprises:

if not, adding one to the accumulated times;

In some embodiments, the method further comprises: if satisfied, no action is taken.

In some embodiments, the method further comprises: and when the accumulated times are not more than the preset times, no operation is performed.

The preset times are not limited in the embodiment of the present application, and may be, for example, 3 times, 5 times, 10 times, 100 times, etc.

The preset personnel in the embodiment of the application can be, for example, the personnel who are appointed to be responsible for updating the similarity model or the preset text.

The second terminal device is not limited in the embodiment of the present application, and may be, for example, an intelligent terminal device such as a mobile phone, a tablet computer, a notebook computer, a desktop computer, an intelligent wearable device, or the second terminal device may be a workstation or a console.

In some embodiments, the identifying the emotion classification of the user according to the feedback information includes: and inputting the feedback information into an emotion recognition model to recognize and obtain the emotion type of the user. The emotion recognition model may be, for example, a deep learning based emotion recognition model.

The emotion classification is not limited by the embodiment of the application, and may include happiness, sadness, anger, fear, surprise, anxiety, aversion, neutrality, and the like, for example.

In some embodiments, the detecting whether the user is satisfied with the service according to the emotion classification of the user includes: obtaining a corresponding relation between emotion types and satisfaction evaluation results, wherein the satisfaction evaluation results are used for indicating satisfaction or dissatisfaction; acquiring a satisfaction evaluation result corresponding to the emotion category of the user according to the emotion category of the user and the corresponding relation; and according to the satisfaction evaluation result corresponding to the emotion type of the user, confirming that the user is satisfied with the service or is not satisfied with the service.

That is, a correspondence relationship between the emotion type and the satisfaction evaluation result may be established in advance. For example, when the emotion classification is happy, the corresponding satisfaction evaluation result is satisfied; when the emotion classification is anxiety, the corresponding satisfaction evaluation result is dissatisfaction. And then, judging whether the user is satisfied with the service according to the satisfaction evaluation result corresponding to the emotion type of the user. If the satisfaction evaluation result of the user is satisfied, confirming that the user is satisfied with the service; if the satisfaction evaluation result of the user is unsatisfactory, the user is confirmed to be unsatisfactory for the service.

In the embodiment of the application, the service items correspond to preset texts and preset actions, and the first interactive video corresponding to the service items in the question-answer library is synthesized according to the preset texts and the preset actions of the service items. Specifically, the synthesis process may be, for example: acquiring a preset text and a preset action corresponding to the service item; and performing voice synthesis according to the preset text to obtain voice data, driving the mouth shape and expression of the virtual service personnel according to the voice data to obtain face driving data, driving the motion of the virtual service personnel according to the preset motion to obtain motion driving data, and generating a first interactive video corresponding to the service item according to the voice data, the face driving data and the motion driving data. In the first interactive video, the mouth shape, expression and action of the virtual attendant can be changed along with the interaction process so as to be close to the real attendant.

Therefore, the satisfaction degree of the user on the service provided by the virtual service personnel can be better known by receiving the feedback information of the user and identifying the emotion type of the user, and the service quality can be further optimized. According to the emotion type of the user, whether the user is satisfied with the service or not is detected, and service matters with problems can be found in time, so that the method is improved in a targeted manner. When the accumulated times are greater than the preset times, prompt information is sent to a second terminal device of the preset personnel, so that the related personnel can be reminded of updating the similarity model or adjusting the preset text, and the matching accuracy and the user satisfaction of the virtual service personnel are improved. The method strengthens the accuracy and reliability of the virtual service personnel when processing the user demands, thereby improving the user satisfaction degree and the office efficiency and enabling enterprises and government departments to better meet the client demands. Through user feedback and emotion recognition, the technical scheme is beneficial to collecting precious user opinions and suggestions, provides powerful support for further optimizing the working effect of virtual service personnel, ensures that related personnel pay attention to the problem of user satisfaction in time, and prevents the quality of service from sliding down. In conclusion, the technical scheme can continuously optimize the service quality and the matching accuracy of the virtual service personnel by collecting the user feedback information, identifying the emotion type of the user and sending the prompt information.

Illustrating the user interaction with the virtual attendant "small a".

The user may provide feedback information to the virtual attendant "small a" via text, voice, or images, such as the user entering a piece of voice information: "the business registration process you just give is not clear. "in this example, the target service transaction is an enterprise registration.

Identifying the emotion category of the user by analyzing the feedback information of the user, such as: "dissatisfaction".

And judging whether the user is satisfied with the service according to the emotion type of the user. In this example, the user is not satisfied with the service.

The accumulated number of times is incremented by one whenever the user is not satisfied with the service. For example, the current cumulative number of times is 3. The preset number of times is, for example, 5 times, as a threshold for user dissatisfaction with the service.

When the accumulated times are greater than the preset times (for example, the accumulated times reach 6 times), a prompt message is sent to a mobile phone (namely, a second terminal device) of a preset person, and the content is for example: "receiving multiple unsatisfactory feedback about enterprise registration service events, please check and update the similarity model or preset text. "wherein the preset text is a standard reply used by the virtual attendant" small a "in answering user questions, such as a description of the enterprise registration process.

In some embodiments, the method further comprises:

Therefore, through displaying a plurality of second interactive videos related to the first interactive video on the first terminal equipment, richer and more comprehensive service information can be provided for the user, and the user is helped to better know the notice matters, common problems or related service matters of the target service matters. The user can select to watch the second interactive video related to the target service item according to the own requirements and interests, so that the initiative and the interactivity of the user in the process of acquiring the service information are improved, and the user satisfaction is improved. The method provides a more convenient and personalized information acquisition mode for the user, and is beneficial to meeting the requirements and preferences of different users.

Illustrating the user interaction with the virtual attendant "small a".

The user accesses the government virtual attendant "small a" by a browser or a dedicated government service application (e.g., a "government service") using a smart phone or computer (i.e., a first terminal device). In this example, the target service transaction is an enterprise registration.

The virtual servicer 'small A' plays a first interactive video introducing the enterprise registration process and the office manual according to the enterprise registration problem presented by the user.

After playing the first interactive video, the first terminal device displays a plurality of second interactive videos related to enterprise registration, for example: "Material required for Enterprise registration", "tax handling for Enterprise registration", and "common questions for Enterprise registration".

Notice of enterprise registration, such as valid credentials to be submitted during registration, applicant's age restrictions, etc. Common issues with enterprise registration, such as enterprise type selection issues, registration capital requirement issues, registration address issues, enterprise name specification issues, legal representative identity restriction issues, and the like. The relevant service matters registered by the enterprises are tax registration, social insurance registration and the like.

The user clicks on one of the second interactive videos, such as "materials required for enterprise registration", and the first terminal device plays the video to detail various materials and submission requirements required for enterprise registration.

Referring to fig. 4, fig. 4 is a schematic flow chart of a downloading certification document according to an embodiment of the present application.

In some embodiments, in the first interactive video, the virtual attendant directs the user to click a document download button to download a certification document;

the method further comprises the steps of:

The embodiment of the application does not limit the type of the certification document, and can be, for example, birth certificates, household books, identity certificates, wedding certificates, divorce certificates, academic certificates, employment certificates, social security certificates, real estate certificates and the like.

Therefore, in the first interactive video, the virtual service personnel guide the user to click the document downloading button, so that the user can easily acquire the required proving document, and the user experience and satisfaction degree are improved. By responding to the clicking operation of the user, the document is downloaded to the first terminal device, so that convenience and high efficiency of the document downloading function are realized, and time and energy of the user are saved. The technical scheme simplifies the acquisition process of the proving document, thereby improving the working efficiency. In the interaction process of the virtual service personnel, the downloading function of the proof document is realized, the service content provided by the virtual service personnel is further enriched, and the user can complete more service matters on one platform.

By way of example, assuming a user wants to know how to download an academic proof document, a particular website or mobile application may be accessed using a smartphone or computer. On a website or mobile application, a first interactive video of a virtual attendant "small a" is played. In the first interactive video, the small a introduces the use of the academic proof document, such as job hunting, job improvement, reservation, etc., and how to acquire the academic proof document.

In the first interactive video, the small A guides the user to click a 'document download button' on the page, and tells the user to click the button, and then automatically downloads the academic proof document. User click operation: the user clicks the "document download button" as directed by small a. And downloading the academic proof document to the first terminal equipment of the user according to the clicking operation of the user. Through the interaction mode, a user can conveniently acquire the academic evidence document, and the working efficiency is improved. Meanwhile, the small A of the virtual service personnel provides detailed guidance in the whole process, so that the user can obtain the academic proof document more clearly and efficiently.

With continued reference to fig. 4, in some embodiments, when a click operation for the document download button is received, downloading a proof document corresponding to the document download button to the first terminal device includes:

In the embodiment of the application, the identity information to be identified can be, for example, an account number and a password to be identified, fingerprint information, face information, voiceprint information, iris information and the like. Correspondingly, the authenticated identity information is an authenticated account number, a password, fingerprint information, face information, voiceprint information, iris information and the like.

Therefore, after clicking operation for the document downloading button is received, the document downloading interface is displayed, and the document downloading interface is provided with the identity recognition button for triggering the identity authentication step of the user, so that a safe identity authentication mechanism is provided, and the downloading safety and privacy protection of the proving document are ensured. After clicking the identity recognition button, the first terminal device acquires the identity information to be recognized of the user, and is helpful to ensure that the certification document acquired by the user accords with the identity of the user. By detecting the matching condition of the identity information to be identified and the authenticated identity information, the malicious user can be prevented from downloading the proving document of other people to a certain extent, and the information security is improved. If the identity information to be identified is matched with the authenticated identity information, the proving document is downloaded to the first terminal equipment, so that a user can successfully acquire the required proving document, and the user experience and satisfaction are improved.

With continued reference to fig. 4, in some embodiments, said downloading said attestation document to said first terminal device if there is a match comprises:

and downloading the proving document to the first terminal equipment.

And the authentication information is used for inquiring the user information database corresponding to the target service item according to the authenticated identity information, generating a proving document according to the inquiring result, ensuring that the generated proving document is based on the real information of the user, and ensuring the timeliness, the reliability and the accuracy of the generated proving document. In the interaction process of the virtual service personnel, the generation and downloading functions of the certification document are realized, and the service content provided by the virtual service personnel is further enriched.

By way of example, assuming a user wants to download his own social security clearance, a smart phone or computer may be used to access a government website or mobile application. On a website or mobile application, a user enters a piece of text information: "buying a house requires social security. The user's target service item is a social security download.

The first terminal device plays a first interactive video related to a social security plaintext file, wherein, a small A describes in detail how to download the social security certification document. After the first interactive video is played, the user clicks a 'document downloading button' on the page, and jumps to a document downloading interface from the current interface, wherein the document downloading interface is provided with an 'identity recognition button'. The user clicks an identity recognition button, and the first terminal device obtains identity information to be recognized of the user, such as face information to be recognized. And detecting whether the identity information to be identified of the user is matched with authenticated identity information (such as face information submitted when registering or authenticating on websites or mobile applications of various government departments) of the user. If the identity information to be identified is matched with the authenticated identity information, inquiring a user information database corresponding to the social security plaintext file according to the authenticated identity information, generating a social security certification document according to the inquiring result, and downloading the generated social security plaintext file to the first terminal equipment of the user.

In some embodiments, the method further comprises:

In the embodiment of the application, the authenticated identity information of the user can also comprise a name, an identity card number, an age, a occupation, a region where the user is located, a work unit, a work address, a position, a job, a home address, an academic calendar and the like.

In the embodiment of the present application, the interactive data may include, for example, input information (such as text, voice, and image) of the user, operation records (clicking, sliding, etc.), browsing records (page access, stay time, etc.), feedback information (satisfaction, advice, etc.), personal information (authenticated identity information), etc. These data help to improve the quality of service, personalized recommendations, and user experience for the virtual attendant.

Therefore, by combining authenticated identity information of the user and interaction data between the user and the virtual service personnel, one or more recommendation policies corresponding to the user can be acquired more accurately, and therefore pertinence and accuracy of recommendation are improved. The recommended policies are displayed on the first terminal device in a list form, so that a clear and visual information display mode is provided for the user, and the user can quickly browse and know the policies related to the user. The method realizes the policy recommendation function, can help the user to know the policy information related to the user in time, improves the attention degree and the cognition degree of the user on the policy, is beneficial to the benefit of the user from the policy, and avoids the user from missing the policy benefit related to the user. By recommending the policies related to the identity and the requirements of the user, the policy coverage and implementation effect can be improved, and the overall benefit of policy implementation can be improved.

Assuming that the user is a primary enterprise, the user's interaction with the government virtual attendant "small A" will be illustrated.

The user uses a smart phone or a computer (i.e. a first terminal device) to access a website or a mobile application of a government agency, interact with a virtual attendant "small a", and inquire about information about a startup support policy, tax offers and the like.

And analyzing policies possibly interested by the user, such as a business support policy, tax preference policy, talent training policy and the like of the region where the user is located according to authenticated identity information (such as name, identity card number, age, occupation, region where the user is located and the like) and interaction data (such as questions and keywords inquired by the user).

And arranging the entrepreneur support policy, tax preference policy, talent training policy and the like of the region where the user is located into a list form, and displaying the list form on the first terminal equipment of the user.

For example, the recommended policy list may include:

entrepreneur support policy: government policies such as funding support, site offers, etc. for the original enterprise.

Tax preference policy: tax deduction policy for the original enterprise is reduced, so that the operation cost of the enterprise is reduced.

Talent training policy: government provides specialized training and skill improvement courses for the original enterprise.

The method for obtaining the semantic extraction model and the similarity model is not limited in the embodiment of the application, and in some embodiments, the models can be obtained through training, and in other embodiments, the models can be obtained through training in advance.

When each model is obtained through training in a deep learning mode, a proper amount of neuron computing nodes and a multi-layer operation hierarchical structure are established through design, a proper input layer and a proper output layer are selected, a preset deep learning model corresponding to each model (namely an initial model corresponding to each model) can be obtained, a functional relation from input to output is established through learning and optimization of the deep learning model, although the functional relation from input to output cannot be found 100%, the functional relation can be as close to the actual association relation as possible, and therefore each model obtained through training can obtain corresponding output data based on input data.

Training the deep learning model by using training sets corresponding to the models, quickly modeling can be performed by learning a small number of samples, training errors of the deep learning model can be gradually reduced in the continuous training process, and the optimal weight is stored and read; recording the accuracy of the training set and the verification set, and facilitating parameter adjustment (adjustment of model parameters); the model parameters of the deep learning model are updated, so that the model can be better fitted with data, the generalization capability is effectively achieved, and the robustness and the fitting precision are improved.

In some alternative embodiments, the historical data may be data mined to obtain sample data in the training set. That is, the sample data may be collected during the actual interaction of the user with the human attendant (i.e., the human attendant). In addition, the sample data may be automatically generated by using a GAN model generation network.

The GAN model generates an countermeasure network (Generative Adversarial Network) composed of a generation network and a discrimination network. The generation network samples randomly from the potential space (latency space) as input, the output of which needs to mimic as much as possible the real samples in the training set. The input of the discrimination network is then the real sample or the output of the generation network, the purpose of which is to distinguish the output of the generation network as far as possible from the real sample. And the generation of the network should be as fraudulent as possible to discriminate the network. The two networks are mutually opposed and continuously adjust parameters, and the final purpose is that the judging network can not judge whether the output result of the generated network is real or not. A large amount of sample data can be generated by using the GAN model and used for the training process of each model, so that the data volume of original data acquisition can be effectively reduced, and the cost of data acquisition and labeling is greatly reduced.

The training process of each model is not limited, and for example, a training mode of supervised learning, a training mode of semi-supervised learning or a training mode of unsupervised learning can be adopted.

When a training mode of supervised learning or semi-supervised learning is adopted, the method for acquiring the annotation data is not limited, and for example, a manual annotation mode or an automatic annotation or semi-automatic annotation mode can be adopted. When the sample data is acquired in the real interaction process, the real data can be acquired from the historical data in a keyword extraction mode to serve as the annotation data.

The training ending condition in the training process of each model is not limited in the embodiment of the application, for example, the training times can reach the preset times (the preset times are, for example, 1 time, 3 times, 10 times, 100 times, 1000 times, 10000 times, etc.), or the training data in the training set can be all trained once or more times, or the total loss value obtained in the training is not greater than the preset loss value.

In a specific application scenario, the embodiment of the application further provides a virtual attendant interaction method, which comprises the following steps:

when the highest similarity is larger than a preset similarity, taking the service item corresponding to the highest similarity as the target service item;

when the highest similarity is not greater than the preset similarity, playing a second interactive video by using the first terminal equipment so as to guide the user to click a form button and fill in a form; when a clicking operation for the form button is received, displaying the form by using the first terminal equipment, wherein the form is provided with a service item input box and/or a service item drop-down menu; acquiring target service items of the user according to the filling result of the form;

using the first terminal equipment to play the first interactive video, wherein in the first interactive video, the virtual service personnel introduces a business process and/or a business guide corresponding to the target service item; in the first interactive video, the virtual service personnel guide the user to click a document download button to download a proving document;

receiving feedback information of the user by using the first terminal equipment;

if satisfied, do nothing;

if not, adding one to the accumulated times;

when the accumulated times are greater than preset times, carrying out zero clearing processing on the accumulated times, and sending prompt information to a second terminal device of a preset person corresponding to the target service item so as to prompt the preset person to update the similarity model or update a preset text corresponding to the target service item;

When the accumulated times are not more than the preset times, no operation is performed;

when a clicking operation aiming at one of the second interactive videos is received, playing the clicked second interactive video by using the first terminal equipment;

if the two types of data are not matched, no operation is performed;

Downloading the certification document to the first terminal device;

The process of obtaining the similarity between the semantic information and each service item comprises the following steps:

for each of the service matters, the following processing is performed:

the training process of the similarity model comprises the following steps:

for each of the training data, the following processing is performed:

Assuming that the user MING wants to know about the flow of transacting a driver's license change, his interaction with the virtual attendant "small A" of the government agency is illustrated.

MING uses the mobile phone to access government website or mobile application and send access request. The first terminal equipment (mobile phone) establishes communication connection with a target server, and the target server provides a virtual service personnel interaction function.

MING inputs text information: "how to handle the license change? "the handset sends the input information to the server.

The server extracts semantic information from the input information using a semantic extraction model.

The server calculates the similarity between the input semantic information and each service item, and finds the service item corresponding to the highest similarity, namely ' driver's license change card '.

The server acquires a first interactive video corresponding to the driving license change certificate from a question and answer library of a virtual attendant 'small A'.

The mobile phone of MING plays the first interactive video. In the first interactive video, the virtual attendant "small a" introduced the flow and guide for handling driver's license changes, directing the ping to click on the document download button to download the certification document.

The MING clicks a document download button in the first interactive video. The mobile phone displays a document downloading interface and is provided with an identity recognition button.

The MING clicks the identity recognition button, and the mobile phone acquires the identity information to be recognized of the MING.

The server detects whether the identity information to be identified is matched with the authenticated identity information.

After the matching is detected, the server queries a user information database corresponding to the license changing service item of the driving license according to the authenticated identity information, and generates a proving document.

The document was certified for download to the mobile phone of the MING.

According to the authenticated identity information of the MING and the interaction data of the virtual service personnel, the server acquires one or more recommended policies corresponding to the MING, including a new energy automobile preferential purchase policy, a driver safety education and training policy, a traffic jam relieving policy, an insurance preferential policy, a holiday passing preferential policy and the like, and displays the recommended policies on the mobile phone in a list form.

It can be seen that the user MING easily obtains the information and the proof document required for transacting the driver's license change through the interaction with the virtual serviceman "small A", and also obtains the recommendation policy related to the user himself.

Furthermore, in the virtual service personnel interaction method, a personalized recommendation function can be added, so that more relevant service item suggestions are provided for the user. In determining the target service event, the user's historical interaction records, interests, and needs are taken into account to provide personalized service event recommendations that more closely match the user's needs. When determining service matters according to user input information, a multi-mode matching method can be introduced, text, voice and image information input by a user are comprehensively considered, and the matching accuracy of the service matters is improved. When the similarity between the semantic information and the service matters is obtained, a transfer learning method can be adopted, and a pre-trained deep learning model is utilized for rapid training and optimization so as to improve the training efficiency and the matching accuracy of the model. In the user feedback link, an image acquisition component can be used for acquiring face images of the user in real time, and the emotion type of the user is identified and the satisfaction degree of the user is monitored by combining feedback information and the face images of the user. When the user dissatisfaction is detected, the method can actively provide problem solutions or suggestions, timely adjust service strategies and improve user experience. When the first interactive video and the second interactive video are played, a self-adaptive streaming media technology can be adopted, video definition can be automatically adjusted according to the network environment and the equipment performance of the user, and viewing experience of the user under different network environments is guaranteed. In the step of guiding the user to download the proof document, a digital signature and encryption function can be added to ensure the security and the integrity of the document downloaded by the user. In the identification link, various identification modes such as face recognition, fingerprint recognition and the like can be supported, and convenience and safety of user verification are improved. In the step of generating the proving document, the proving document with various formats, such as PDF, word and the like, can be provided according to the requirements and preferences of users, so that the users can conveniently download the required format types.

According to the embodiment of the application, when the virtual service personnel guides the user to operate, the first interactive video aiming at different versions of different users can be provided according to the operation habit, skill level and requirements of the user, so that more intelligent and personalized guidance is realized, and the user experience is improved. For example, for elderly users and vision impaired users, the first interactive video has a larger font size and a thicker font; for a child user, the first interactive video is happy to calm the child's emotion and to draw a distance from the child.

(electronic device)

The embodiment of the application also provides an electronic device, the specific embodiment of which is consistent with the embodiment described in the method embodiment and the achieved technical effect, and part of the contents are not repeated.

In some embodiments, the at least one processor is configured to determine one of a plurality of service items as a target service item for the user from the input information when executing the computer program in the following manner:

In some embodiments, the at least one processor is configured to execute the computer program to determine one of a plurality of service items as a target service item for the user from the input information further in the following manner:

In some embodiments, the at least one processor is configured to obtain the similarity of the semantic information to each of the service matters when executing the computer program by:

for each of the service matters, the following processing is performed:

the training process of the similarity model comprises the following steps:

for each of the training data, the following processing is performed:

In some embodiments, the at least one processor is configured to execute the computer program to further implement the steps of:

if not, adding one to the accumulated times;

the at least one processor is configured to execute the computer program to further implement the steps of:

In some embodiments, the at least one processor is configured to download the document corresponding to the document download button to the first terminal device when executing the computer program in the following manner:

In some embodiments, the at least one processor is configured to download the attestation document to the first terminal device when executing the computer program in the following manner:

and downloading the proving document to the first terminal equipment.

Referring to fig. 5, fig. 5 is a block diagram of an electronic device 10 according to an embodiment of the present application.

The electronic device 10 may for example comprise at least one memory 11, at least one processor 12 and a bus 13 connecting the different platform systems.

Memory 11 may include (computer) readable media in the form of volatile memory, such as Random Access Memory (RAM) 111 and/or cache memory 112, and may further include Read Only Memory (ROM) 113.

The memory 11 also stores a computer program executable by the processor 12 to cause the processor 12 to implement the steps of any of the methods described above.

Memory 11 may also include utility 114 having at least one program module 115, such program modules 115 include, but are not limited to: an operating system, one or more application programs, other program modules, and program data, each or some combination of which may include an implementation of a network environment.

Accordingly, the processor 12 may execute the computer programs described above, as well as may execute the utility 114.

The processor 12 may employ one or more application specific integrated circuits (ASICs, application Specific Integrated Circuit), DSPs, programmable logic devices (PLD, programmableLogic devices), complex programmable logic devices (CPLDs, complex Programmable Logic Device), field programmable gate arrays (FPGAs, fields-Programmable Gate Array), or other electronic components.

Bus 13 may be a local bus representing one or more of several types of bus structures including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, a processor, or any of a variety of bus architectures.

The electronic device 10 may also communicate with one or more external devices such as a keyboard, pointing device, bluetooth device, etc., as well as one or more devices capable of interacting with the electronic device 10 and/or with any device (e.g., router, modem, etc.) that enables the electronic device 10 to communicate with one or more other computing devices. Such communication may be via the input-output interface 14. Also, the electronic device 10 may communicate with one or more networks such as a Local Area Network (LAN), a Wide Area Network (WAN) and/or a public network, such as the Internet, through a network adapter 15. The network adapter 15 may communicate with other modules of the electronic device 10 via the bus 13. It should be appreciated that although not shown, other hardware and/or software modules may be used in connection with the electronic device 10 in actual applications, including, but not limited to: microcode, device drivers, redundant processors, external disk drive arrays, RAID systems, tape drives, data backup storage platforms, and the like.

(computer-readable storage Medium)

The embodiment of the application also provides a computer readable storage medium, and the specific embodiment of the computer readable storage medium is consistent with the embodiment recorded in the method embodiment and the achieved technical effect, and part of the contents are not repeated.

The computer readable storage medium stores a computer program which, when executed by at least one processor, performs the steps of any of the methods or performs the functions of any of the electronic devices described above.

The computer readable medium may be a computer readable signal medium or a computer readable storage medium. In embodiments of the present application, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. The computer readable storage medium can be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or a combination of any of the foregoing. More specific examples (a non-exhaustive list) of the computer-readable storage medium would include the following: an electrical connection having one or more wires, a portable disk, a hard disk, random Access Memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

The computer readable storage medium may include a data signal propagated in baseband or as part of a carrier wave, with readable program code embodied therein. Such a propagated data signal may take any of a variety of forms, including, but not limited to, electro-magnetic, optical, or any suitable combination of the foregoing. A computer readable storage medium may also be any computer readable medium that can transmit, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code embodied on a computer readable storage medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing. Program code for carrying out operations of the present invention may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, C++ or the like and conventional procedural programming languages, such as the C programming language or similar programming languages. The program code may execute entirely on the user's computing device, partly on the user's device, as a stand-alone software package, partly on the user's computing device, partly on a remote computing device, or entirely on the remote computing device or server. In the case of remote computing devices, the remote computing device may be connected to the user computing device through any kind of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or may be connected to an external computing device (e.g., connected via the Internet using an Internet service provider).

(computer program product)

The embodiment of the application also provides a computer program product, the specific embodiment of which is consistent with the embodiment described in the method embodiment and the achieved technical effect, and part of the contents are not repeated.

The present application provides a computer program product comprising a computer program which, when executed by at least one processor, performs the steps of any of the methods or performs the functions of any of the electronic devices described above.

Referring to fig. 6, fig. 6 is a schematic structural diagram of a computer program product according to an embodiment of the present application.

The computer program product is configured to implement the steps of any of the methods described above or to implement the functions of any of the electronic devices described above. The computer program product may employ a portable compact disc read only memory (CD-ROM) and comprise program code and may run on a terminal device, such as a personal computer. However, the computer program product of the present application is not limited thereto, and the computer program product may employ any combination of one or more computer readable media.

The present application has been described in terms of its purpose, performance, advancement, and novelty, and the like, and is thus adapted to the functional enhancement and use requirements highlighted by the patent statutes, but the description and drawings are not limited to the preferred embodiments of the present application, and therefore, all equivalents and modifications that are included in the construction, apparatus, features, etc. of the present application shall fall within the scope of the present application.

Claims

1. A method of virtual attendant interaction, the method comprising:

2. The virtual attendant interaction method of claim 1, wherein the determining one of a plurality of service items as the target service item of the user based on the input information comprises:

3. The virtual attendant interaction method of claim 2, wherein the determining one of a plurality of service items as the target service item for the user based on the input information, further comprises:

4. The virtual attendant interaction method of claim 2, wherein the process of obtaining the similarity of the semantic information to each of the service items comprises:

For each of the service matters, the following processing is performed:

the training process of the similarity model comprises the following steps:

for each of the training data, the following processing is performed:

5. The virtual attendant interaction method of claim 4, further comprising:

if not, adding one to the accumulated times;

6. The virtual attendant interaction method of claim 1, further comprising:

7. The virtual attendant interaction method of claim 1, wherein in the first interactive video, the virtual attendant directs the user to click a document download button to download a certification document;

the method further comprises the steps of:

8. The virtual serviceman interaction method according to claim 7, wherein the downloading of the certification document corresponding to the document download button to the first terminal device when the click operation for the document download button is received comprises:

9. The virtual attendant interaction method of claim 8, wherein the downloading the certification document to the first terminal device if matched comprises:

and downloading the proving document to the first terminal equipment.

10. The virtual attendant interaction method of claim 9, further comprising:

11. An electronic device comprising a memory and at least one processor, the memory storing a computer program, the at least one processor being configured to implement the following steps when executing the computer program:

12. A computer-readable storage medium, characterized in that it stores a computer program which, when executed by at least one processor, implements the steps of the method of any one of claims 1-10 or the functions of the electronic device of claim 11.

13. A computer program product, characterized in that it comprises a computer program which, when executed by at least one processor, implements the steps of the method of any one of claims 1-10 or the functions of the electronic device of claim 11.