Method and device for implementing interaction optimization applied to digital human

Abstract

The application relates to the technical field of computers and discloses an interactive optimization implementation method and device applied to digital people, which comprises the following steps: collecting interactive records corresponding to a plurality of digital people in a target time period, analyzing each interactive record, obtaining an interactive control strategy to be improved, determining an interactive optimization direction corresponding to the interactive control strategy to be improved, and performing an optimization operation on the interactive control strategy to be improved according to the interactive optimization direction to update the interactive control strategy to be improved. It can be seen that the application can improve the interactive experience of users, improve the interactive quality of digital people, improve the service experience and effect of digital people, and simplify the interactive process to improve the interactive efficiency of digital people.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer, in particular to an interactive optimization implementation method and device applied to digital person. BACKGROUND

[0002] In the current rapid development of science and technology, the maturity of artificial intelligence and virtual reality technology brings vitality to the virtual digital person industry. The production of digital person is simplified, the technical availability is significantly improved, the intelligent driving of text, voice and video can be realized, and the intelligent automatic broadcasting can be realized through AI and NLP technology to meet the personalized needs of users, which greatly promotes the business development.

[0003] However, the development of virtual digital person in the meta universe is still in its infancy. In terms of digital person interaction, the existing interaction control strategy has many shortcomings. On the one hand, part of the interaction strategy cannot fully improve the user experience and cannot accurately grasp the user's intention, resulting in a stiff and unsmooth interaction process; on the other hand, some interaction strategies can meet the basic experience, but the process is complicated and lacks further simplification and optimization. Therefore, it is particularly important to propose a technical solution that can improve the quality of digital person interaction and improve the service experience of digital person. SUMMARY

[0004] The present application provides an interactive optimization implementation method and device applied to digital person, which can improve the quality of digital person interaction and improve the service experience of digital person.

[0005] In order to solve the above technical problems, the first aspect of the present application discloses an interactive optimization implementation method applied to digital person, which comprises:

[0006] Collecting the interaction records corresponding to a plurality of digital persons in a target time period, and analyzing each of the interaction records to obtain an interaction control strategy to be improved;

[0007] Determining the interaction optimization direction corresponding to the interaction control strategy to be improved, and performing optimization operation on the interaction control strategy to be improved according to the interaction optimization direction to update the interaction control strategy to be improved.

[0008] As an optional implementation manner, in the first aspect of the present application, the analysis of each of the interaction records to obtain the interaction control strategy to be improved comprises:

[0009] For each of the interaction records, the interaction purpose of the target user in the interaction record and the interaction control strategy are determined;

[0010] For each of the interaction records, analyze the interaction control strategy of the interaction record, determine interaction step information corresponding to the interaction control strategy, and the interaction step information includes at least one interaction step and interaction time corresponding to each of the interaction steps;

[0011] For each of the interaction records, determine the implementation step information corresponding to the implementation of the interaction purpose in the interaction record according to the interaction step information of the interaction record and the interaction purpose of the target user;

[0012] For each of the interaction records, determine whether the implementation step information satisfies the preset step optimal condition, and when the implementation step information does not satisfy the preset step optimal condition, determine that the interaction control strategy of the interaction record is the interaction control strategy to be improved.

[0013] As an optional implementation, in the first aspect of the application, the implementation step information includes at least one implementation step corresponding to the implementation of the interaction purpose and the implementation time of each of the implementation steps;

[0014] For each of the interaction records, determine whether the implementation step information satisfies the preset step optimal condition, and when the implementation step information does not satisfy the preset step optimal condition, determine that the interaction control strategy of the interaction record is the interaction control strategy to be improved.

[0015] For each of the interaction records, obtain the interaction requirement information corresponding to the target user in the interaction record, and the interaction requirement information includes interaction step requirements and interaction time requirements;

[0016] For each of the interaction records, determine whether the step number of the implementation step satisfies the interaction step requirement and whether the implementation time of each of the implementation steps satisfies the interaction time requirement, and when the step number satisfies the interaction step requirement and the implementation time of each of the implementation steps satisfies the interaction time requirement, analyze the key information corresponding to the interaction purpose of the target user;

[0017] For each of the interaction records, determine the first step number of the interaction step in which the key information is obtained and the second step number of the interaction step in which the interaction purpose is implemented, and determine whether the number difference between the first step number and the second step number is greater than a preset number difference threshold, and when the number difference is greater than the number difference threshold, determine that the implementation step information does not satisfy the preset step optimal condition.

[0018] As an optional implementation, in the first aspect of the application, the interaction optimization direction includes a first interaction optimization direction, and further includes a second interaction optimization direction or a third interaction optimization direction;

[0019] The determination of the interaction optimization direction corresponding to the interaction control strategy to be improved includes:

[0020] determining at least one type of interaction user and interaction habits of each type of the interaction user according to the interaction control strategy to be improved;

[0021] For each type of the interaction user, determining a matching degree between each interaction step and the interaction habits of the interaction user according to interaction step information of the interaction control strategy to be improved, and determining whether each matching degree is greater than a preset matching degree threshold;

[0022] determining an interaction step with a matching degree less than the matching degree threshold as a target interaction step, and determining a first interaction optimization direction corresponding to the interaction control strategy to be improved according to each target interaction step, the first interaction optimization direction including a first interaction step optimization direction;

[0023] when the number of steps of the interaction control strategy to be improved does not meet the interaction step requirement, or the implementation time length of at least one of the implementation steps does not meet the interaction time length requirement, determining a second interaction optimization direction corresponding to the interaction control strategy to be improved, the second interaction optimization direction including a second interaction step optimization direction and / or an interaction time length optimization direction;

[0024] when the number difference of the interaction control strategy to be improved is greater than the number difference threshold, determining a third interaction optimization direction corresponding to the interaction control strategy to be improved according to the number difference, the third interaction optimization direction including a third interaction step optimization direction.

[0025] As an optional implementation, in the first aspect of the present application, the method further comprises:

[0026] determining deviation information between an interaction result of the interaction control strategy to be improved and an interaction purpose of a target user corresponding to the interaction control strategy to be improved;

[0027] For each interaction step of the interaction control strategy to be improved, determining whether the target user of the interaction control strategy to be improved feeds back user negative feedback information in the interaction step to obtain an analysis result, the user negative feedback information including at least one of negative feedback voice, negative feedback text, negative feedback action and negative feedback emotion;

[0028] For each interaction step of the interaction control strategy to be improved, when the analysis result indicates that the target user feeds back the user negative feedback information in the interaction step, determining an influence degree of the user negative feedback information on the deviation information;

[0029] And the determining the interaction optimization direction corresponding to the interaction control strategy to be improved comprises:

[0030] The interaction optimization direction corresponding to the interaction control strategy to be improved is determined according to each of the influence degrees.

[0031] As an optional implementation, in the first aspect of the application, the interaction optimization direction comprises a fourth interaction optimization direction.

[0032] The determining the interaction optimization direction corresponding to the interaction control strategy to be improved comprises:

[0033] In each of the interaction steps of the interaction control strategy to be improved, at least one prediction step of predicting the user intention of the digital human is screened, and prediction information of each of the prediction steps is determined.

[0034] For each of the prediction steps, user satisfaction information of the target user in at least one interaction step after the prediction step is obtained, and prediction accuracy of the prediction step is determined according to the user satisfaction information and the prediction information of the prediction step.

[0035] The fourth interaction optimization direction corresponding to the interaction control strategy to be improved is determined according to the prediction accuracy of each of the prediction steps.

[0036] As an optional implementation, in the first aspect of the application, the performing the optimization operation on the interaction control strategy to be improved according to the interaction optimization direction to update the interaction control strategy to be improved comprises:

[0037] The optimization operation is performed on the interaction control strategy to be improved according to the interaction optimization direction to update the interaction control strategy to be improved.

[0038] A current network state is determined, and the current network state comprises an online network state or an offline network state.

[0039] When the current network state comprises the online network state, the updated interaction control strategy to be improved is updated to each digital human terminal through a network, and when the current network state comprises the offline network state, the updated interaction control strategy to be improved is stored into a local digital human database.

[0040] The second aspect of the application discloses an interaction optimization implementation device applied to a digital human, and the device comprises:

[0041] A collection module is configured to collect interaction records corresponding to a plurality of digital humans in a target time period.

[0042] an analysis module, configured to analyze each of the interaction records to obtain an interaction control strategy to be improved;

[0043] a first determination module, configured to determine an interaction optimization direction corresponding to the interaction control strategy to be improved;

[0044] an optimization module, configured to perform an optimization operation on the interaction control strategy to be improved according to the interaction optimization direction, so as to update the interaction control strategy to be improved.

[0045] As an optional implementation, in the second aspect, the manner in which the analysis module analyzes each of the interaction records to obtain an interaction control strategy to be improved specifically includes:

[0046] for each of the interaction records, determining an interaction purpose of the target user and an interaction control strategy in the interaction record;

[0047] for each of the interaction records, analyzing the interaction control strategy in the interaction record to determine interaction step information corresponding to the interaction control strategy, the interaction step information including at least one interaction step and an interaction time length corresponding to each of the interaction steps;

[0048] for each of the interaction records, determining, according to the interaction step information of the interaction record and the interaction purpose of the target user, implementation step information corresponding to achieving the interaction purpose in the interaction record;

[0049] for each of the interaction records, judging whether the implementation step information satisfies a preset step optimization condition, and determining that the interaction control strategy in the interaction record is the interaction control strategy to be improved when the implementation step information does not satisfy the preset step optimization condition.

[0050] As an optional implementation, in the second aspect, the implementation step information includes at least one implementation step corresponding to achieving the interaction purpose and an implementation time length of each of the implementation steps;

[0051] the manner in which the analysis module judges, for each of the interaction records, whether the implementation step information satisfies a preset step optimization condition specifically includes:

[0052] for each of the interaction records, obtaining interaction requirement information corresponding to the target user in the interaction record, the interaction requirement information including interaction step requirements and an interaction time length requirement;

[0053] For each of the interaction records, it is judged whether the number of implementation steps meets the interaction step requirement and whether the implementation duration of each of the implementation steps meets the interaction duration requirement, and when the number of steps meets the interaction step requirement and the implementation duration of each of the implementation steps meets the interaction duration requirement, the key information corresponding to the interaction purpose of the target user is analyzed;

[0054] For each of the interaction records, the first step number of the interaction step in which the key information is obtained and the second step number of the interaction step in which the interaction purpose is implemented are determined, and it is judged whether the number difference between the first step number and the second step number is greater than a preset number difference threshold, and when the number difference is greater than the number difference threshold, it is determined that the implementation step information does not meet the preset step optimization condition.

[0055] As an optional implementation, in the second aspect of the application, the interaction optimization direction includes a first interaction optimization direction, and further includes a second interaction optimization direction or a third interaction optimization direction.

[0056] The first determining module determines the interaction optimization direction corresponding to the interaction control strategy to be improved in the following manner:

[0057] According to the interaction control strategy to be improved, at least one type of interaction user and the interaction habit corresponding to each type of interaction user are determined.

[0058] For each type of interaction user, according to the interaction step information of the interaction control strategy to be improved, the matching degree between each of the interaction steps and the interaction habit of the type of interaction user is determined, and it is judged whether each of the matching degrees is greater than a preset matching degree threshold.

[0059] The interaction step whose matching degree is less than the matching degree threshold is determined as a target interaction step, and according to each of the target interaction steps, a first interaction optimization direction corresponding to the interaction control strategy to be improved is determined, and the first interaction optimization direction includes a first interaction step optimization direction.

[0060] When the number of steps of the interaction control strategy to be improved does not meet the interaction step requirement, or the implementation duration of at least one of the implementation steps does not meet the interaction duration requirement, a second interaction optimization direction corresponding to the interaction control strategy to be improved is determined, and the second interaction optimization direction includes a second interaction step optimization direction and / or an interaction duration optimization direction.

[0061] When the number difference of the to-be-improved interaction control strategy is greater than the number difference threshold, a third interaction optimization direction corresponding to the to-be-improved interaction control strategy is determined according to the number difference, and the third interaction optimization direction includes a third interaction step optimization direction.

[0062] As an optional implementation, in the second aspect of the present application, the device further comprises:

[0063] A second determination module is configured to determine deviation information between an interaction result of the to-be-improved interaction control strategy and an interaction purpose of a target user corresponding to the to-be-improved interaction control strategy.

[0064] The second determination module is further configured to, for each interaction step of the to-be-improved interaction control strategy, determine whether the target user of the to-be-improved interaction control strategy feeds back user negative feedback information in the interaction step, to obtain an analysis result, the user negative feedback information including at least one of negative feedback speech, negative feedback text, negative feedback action, and negative feedback emotion.

[0065] The second determination module is further configured to, for each interaction step of the to-be-improved interaction control strategy, when the analysis result indicates that the target user feeds back the user negative feedback information in the interaction step, determine an influence degree of the user negative feedback information on the deviation information.

[0066] The first determination module determines the interaction optimization direction corresponding to the to-be-improved interaction control strategy in the following manner:

[0067] The interaction optimization direction corresponding to the to-be-improved interaction control strategy is determined according to each influence degree.

[0068] As an optional implementation, in the second aspect of the present application, the interaction optimization direction includes a fourth interaction optimization direction.

[0069] The first determination module determines the interaction optimization direction corresponding to the to-be-improved interaction control strategy in the following manner:

[0070] In each interaction step of the to-be-improved interaction control strategy, at least one prediction step of predicting user intention of the digital human is screened, and prediction information of each prediction step is determined.

[0071] For each prediction step, user satisfaction information of the target user in at least one interaction step after the prediction step is obtained, and prediction accuracy of the prediction step is determined according to the user satisfaction information and the prediction information of the prediction step.

[0072] According to the prediction accuracy of each prediction step, a fourth interaction optimization direction corresponding to the interaction control strategy to be improved is determined.

[0073] As an optional implementation, in the second aspect of the present invention, the optimization module performs an optimization operation on the interactive control strategy to be improved according to the interactive optimization direction to update the interactive control strategy to be improved, specifically including:

[0074] performing an optimization operation on the interaction control strategy to be improved according to the interaction optimization direction to update the interaction control strategy to be improved;

[0075] Determine a current network state, where the current network state includes an online network state or an offline network state;

[0076] When the current network state includes the online network state, the updated interactive control strategy to be improved is updated to each digital human terminal through the network; when the current network state includes the offline network state, the updated interactive control strategy to be improved is stored in the local digital human database.

[0077] A third aspect of the present invention discloses another interactive optimization implementation device for digital humans, the device comprising:

[0078] a memory storing executable program code;

[0079] a processor coupled to the memory;

[0080] The processor calls the executable program code stored in the memory to execute the interactive optimization implementation method applied to digital humans disclosed in the first aspect of the present invention.

[0081] The fourth aspect of the present invention discloses a computer storage medium, which stores computer instructions. When the computer instructions are called, they are used to execute the interactive optimization implementation method for digital humans disclosed in the first aspect of the present invention.

[0082] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0083] In this embodiment of the present invention, interaction records corresponding to several digital humans within a target time period are collected and analyzed to determine the interaction control strategy to be improved. The interaction optimization direction corresponding to the interaction control strategy to be improved is then determined. Based on the interaction optimization direction, an optimization operation is then performed on the interaction control strategy to be improved, thereby updating the interaction control strategy to be improved. This demonstrates that the implementation of this invention can enhance the user's interactive experience, improve the quality of digital human interaction, enhance the digital human service experience and effectiveness, and simplify the interaction process to improve digital human interaction efficiency. BRIEF DESCRIPTION OF DRAWINGS

[0084] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the embodiments will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can be obtained without creative effort based on these drawings.

[0085] Figure 1 is a flow diagram of an interactive optimization implementation method applied to a digital person according to an embodiment of the present application;

[0086] Figure 2 is a flow diagram of another interactive optimization implementation method applied to a digital person according to an embodiment of the present application;

[0087] Figure 3 is a structural diagram of an interactive optimization implementation device applied to a digital person according to an embodiment of the present application;

[0088] Figure 4 is a structural diagram of another interactive optimization implementation device applied to a digital person according to an embodiment of the present application;

[0089] Figure 5 is a structural diagram of another interactive optimization implementation device applied to a digital person according to an embodiment of the present application. DETAILED DESCRIPTION

[0090] In order to make the person skilled in the art better understand the present application, the technical solutions in the embodiments of the present application will be described clearly and completely with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of the present application.

[0091] The terms "first", "second", etc. in the specification and claims of the present application and the above drawings are used to distinguish different objects, not to describe a specific order. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, device, product or end including a series of steps or units is not limited to the listed steps or units, but can optionally include steps or units not listed or can optionally include other steps or units inherent to the process, method, product or end.

[0092] Reference to“an embodiment” herein means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the application. The appearances of the phrase“in one embodiment” in various places in the specification are not necessarily referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. As will be apparent to those of ordinary skill in the art, embodiments described herein can be combined with other embodiments.

[0093] The application discloses an interactive optimization implementation method and device applied to digital people, which can improve the interactive experience of users, improve the interactive quality of digital people, improve the service experience and effect of digital people, and simplify the interactive process to improve the interactive efficiency of digital people. The following will be described in detail.

[0094] Embodiment one

[0095] Please refer to Figure 1 , Figure 1 is a flowchart of an interactive optimization implementation method applied to digital people disclosed by the embodiment of the application. Wherein, Figure 1 The interactive optimization implementation method applied to digital people described can be applied to an interactive optimization implementation device applied to digital people. The interactive optimization implementation device applied to digital people can include a smart device for interacting with a user. The smart device can include any one of a user device, a smart server, or a smart platform. The user device can be built-in with a digital person program for interacting with the user. Specifically, the user device includes but is not limited to at least one of a smart phone (Android phone, iOS phone, etc.) with a digital person interaction function, a smart phone number watch, and a smart home device, etc. The smart server includes a local server or a cloud server, and the embodiment of the application does not make any limitation. As shown in Figure 1 The interactive optimization implementation method applied to digital people can include the following operations:

[0096] 101, collect the interactive records corresponding to a plurality of digital people in a target time period, and analyze each interactive record to obtain an interactive control strategy to be improved.

[0097] In the embodiment of the present application, optionally, the target time period can include a time period of any length before the collection time, for example, interaction records in the past week or month before the collection time, the interaction records corresponding to the plurality of digital humans can include one or more interaction records of a plurality of different digital humans, or one or more interaction records of a digital human, the interaction record can include an interaction control strategy and interaction content in the interaction process, the interaction control strategy can include at least one interaction step and an interaction time corresponding to each interaction step, the interaction content can include interaction content corresponding to the digital human and interaction content corresponding to the user, specifically, the interaction content can include one or more combinations of text content, image content, video content, voice content, action content, and control content, and the present application is not limited.

[0098] In the embodiment of the present application, optionally, each interaction record can be analyzed to obtain an interaction control strategy to be improved, the interaction control strategy to be improved can include an interaction control strategy to be improved to improve user experience, and / or an interaction control strategy to be further simplified on the basis of meeting user experience, and the present application is not limited.

[0099] In the embodiment of the present application, optionally, the interaction record corresponding to each digital human can include an interaction record permission of the interaction record, the interaction record permission can be used to limit the permission requirement of viewing and / or obtaining the interaction record, specifically, when viewing and / or obtaining the interaction record, the execution subject needs to be permission verified, when the permission of the execution subject is less than the interaction record permission of the interaction record, the execution subject is not allowed to view and / or obtain the interaction record, and the execution subject needs to apply for permission to obtain the viewing and / or obtaining permission of the interaction record, and the present application is not limited.

[0100] In the embodiment of the present application, optionally, in the process of collecting the interaction records corresponding to the plurality of digital humans in the target time period, all collected interaction records are data encrypted, and the encrypted interaction records are transmitted, the data encryption mode can include symmetric encryption or asymmetric encryption, after receiving the encrypted interaction records, the data is decrypted based on a preset secret key, and the data is data verified and data integrity checked, and after verification, the operation of analyzing each interaction record to obtain the interaction control strategy to be improved is performed.

[0101] 102. Determine the interaction optimization direction corresponding to the interaction control strategy to be improved, and perform optimization operation on the interaction control strategy to be improved according to the interaction optimization direction to update the interaction control strategy to be improved.

[0102] In the embodiments of the present application, optionally, the interaction optimization direction corresponding to the interaction control strategy to be improved can be determined, the interaction optimization direction can include an interaction step optimization direction and / or an interaction time length optimization direction, specifically, the interaction step optimization direction can include deleting one or more interaction steps in the interaction control strategy to be improved, and / or simplifying one or more interaction steps in the interaction control strategy to be improved, and / or adjusting the execution order of one or more interaction steps in the interaction control strategy to be improved, and the interaction time length optimization direction can include shortening the interaction time length of one or more interaction steps in the interaction control strategy to be improved, which is not limited by the present application.

[0103] In the embodiments of the present application, optionally, the optimization operation can be performed on the interaction control strategy to be improved according to the interaction optimization direction, that is, at least one of the step deletion operation, the step simplification operation and the interaction time length shortening operation is performed on the interaction control strategy to be improved to update the interaction control strategy to be improved, and the updated interaction control strategy to be improved is stored, which is not limited by the present application.

[0104] It can be seen that the implementation of the application Figure 1 The application described herein can collect interaction records corresponding to a plurality of digital humans in a target time period, analyze each interaction record, obtain an interaction control strategy to be improved, determine an interaction optimization direction corresponding to the interaction control strategy to be improved, and perform an optimization operation on the interaction control strategy to be improved according to the interaction optimization direction to update the interaction control strategy to be improved, which can improve the user's interaction experience, improve the quality of digital human interaction, improve the service experience and effect of digital humans, and simplify the interaction process to improve the efficiency of digital human interaction.

[0105] In an optional embodiment, performing an optimization operation on the interaction control strategy to be improved according to the interaction optimization direction to update the interaction control strategy to be improved can include the following operations:

[0106] Performing an optimization operation on the interaction control strategy to be improved according to the interaction optimization direction to update the interaction control strategy to be improved;

[0107] Determine the current network state, the current network state including an online network state or an offline network state;

[0108] When the current network state includes an online network state, the updated interaction control strategy to be improved is updated to each digital human terminal through the network, and when the current network state includes an offline network state, the updated interaction control strategy to be improved is stored in the local digital human database.

[0109] In the optional embodiment, after the optimization operation is performed on the to-be-improved interaction control strategy according to the interaction optimization direction to update the to-be-improved interaction control strategy, the updated to-be-improved interaction control strategy can be processed according to the network state of the digital human. Specifically, the current network state can include an online network state or an offline network state. When the current network state includes the online network state, the updated to-be-improved interaction control strategy is updated to each digital human terminal through the network, that is, the updated to-be-improved interaction control strategy is synchronously shared through the network. Specifically, the synchronous sharing can be performed on a webpage, a mobile terminal, or the like. When the current network state includes the offline network state, the updated to-be-improved interaction control strategy is stored in the local digital human database. The embodiment is not limited.

[0110] It can be seen that by implementing the optional embodiment, the optimization operation can be performed on the to-be-improved interaction control strategy according to the interaction optimization direction to update the to-be-improved interaction control strategy, the current network state is determined, when the current network state includes the online network state, the updated to-be-improved interaction control strategy is updated to each digital human terminal through the network, and when the current network state includes the offline network state, the updated to-be-improved interaction control strategy is stored in the local digital human database. The updated strategy can be processed based on the network state, the execution efficiency of the optimized strategy is improved, and then the user's interaction experience is improved, and the digital human interaction quality, the digital human service experience, and the effect are improved.

[0111] Embodiment two

[0112] Please refer to Figure 2 , Figure 2 is a flowchart of an application of the application embodiment to the interaction optimization implementation method of the digital human. Wherein, Figure 2 The application of the application embodiment to the interaction optimization implementation method of the digital human can be applied to the application of the application of the application embodiment to the interaction optimization implementation device of the digital human. The application of the application embodiment to the interaction optimization implementation device of the digital human can include a smart device for interacting with a user. The smart device can include any one of a user device, a smart server, or a smart platform. The user device can be built-in with a digital human program for interacting with a user. Specifically, the user device includes but is not limited to at least one of a smart phone (Android phone, iOS phone, etc.) with a digital human interaction function, a smart phone number watch, and a smart home device, etc. The smart server includes a local server or a cloud server. The application embodiment is not limited. As Figure 2 shown, the application of the application embodiment to the interaction optimization implementation method of the digital human can include the following operations:

[0113] 201, collect the interaction records corresponding to a plurality of digital humans in a target time period.

[0114] 202. For each interaction record, determine the interaction purpose and interaction control strategy of the target user in the interaction record.

[0115] In an embodiment of the present invention, optionally, for each interaction record, the interaction purpose and interaction control strategy of the target user in the interaction record are determined. The target user may include one or more users. When the target user includes multiple users, a main user and multiple auxiliary users may be determined among the multiple users. The interaction purpose of the target user may include one or more combinations of question answering and query purposes, information screening purposes, entertainment purposes, task setting purposes, device control purposes, etc. The interaction control strategy in the interaction record may be an existing strategy stored in the database, or it may be an interaction control strategy generated for the first time in the interaction record, and the present invention does not limit this.

[0116] 203. For each interaction record, analyze the interaction control strategy of the interaction record to determine interaction step information corresponding to the interaction control strategy.

[0117] In an embodiment of the present invention, optionally, the interaction step information corresponding to the interaction control strategy may include at least one interaction step and the interaction duration corresponding to each interaction step, wherein each interaction step may include specific interaction content, such as user input content, digital human feedback content and predicted content, etc., which is not limited by the present invention.

[0118] 204. For each interaction record, determine, based on the interaction step information of the interaction record and the interaction purpose of the target user, implementation step information corresponding to achieving the interaction purpose in the interaction record.

[0119] In an embodiment of the present invention, optionally, the implementation step information corresponding to achieving the interaction purpose in the interaction record may include at least one implementation step corresponding to achieving the interaction purpose and the implementation duration of each implementation step. It may also include the step number of each implementation step, the specific interaction content in each implementation step, etc. For example, when the user's interaction purpose is to answer a question, one round of interaction between the digital human and the user can be defined as an interaction step. When the digital human gives an answer to the question in the third round of interaction, the first three rounds of interaction steps can be determined as implementation steps. The present invention does not limit this.

[0120] 205. For each interaction record, determine whether the implementation step information meets the preset step optimal condition. When the implementation step information does not meet the preset step optimal condition, determine that the interaction control strategy of the interaction record is the interaction control strategy to be improved.

[0121] In the embodiment of the present application, optionally, it can be judged whether the implementation step information meets the preset step optimal condition. The step optimal condition can be used to limit the number and duration of the implementation steps. When the implementation step information does not meet the preset step optimal condition, it can be determined that the interaction control strategy of the interaction record is the interaction control strategy to be improved. The present application does not limit this.

[0122] 206. Determine the interaction optimization direction corresponding to the interaction control strategy to be improved, and perform optimization operation on the interaction control strategy to be improved according to the interaction optimization direction to update the interaction control strategy to be improved.

[0123] In the embodiment of the present application, for other descriptions of steps 201 and 206, please refer to the detailed description of steps 101 and 102 in the first embodiment of the present application. The present application will not be repeated here.

[0124] It can be seen that the implementation Figure 2 The described interaction optimization implementation method applied to digital people can collect interaction records corresponding to a plurality of digital people in a target time period. For each interaction record, the interaction purpose of the target user and the interaction control strategy in the interaction record are determined. The interaction control strategy of the interaction record is analyzed. The interaction step information corresponding to the interaction control strategy is determined. The implementation step information corresponding to the implementation of the interaction purpose in the interaction record is determined according to the interaction step information of the interaction record and the interaction purpose of the target user. When the implementation step information does not meet the preset step optimal condition, it is determined that the interaction control strategy of the interaction record is the interaction control strategy to be improved. The accuracy and reliability of determining the interaction control strategy to be improved can be improved. The pertinence and accuracy of optimizing the strategy are improved. The interaction optimization direction corresponding to the interaction control strategy to be improved is determined. The optimization operation is performed on the interaction control strategy to be improved according to the interaction optimization direction to update the interaction control strategy to be improved. The interaction experience of the user can be improved. The quality of digital human interaction, the service experience and effect of digital human are improved. The interaction process can be simplified to improve the efficiency of digital human interaction.

[0125] In an optional embodiment, the implementation step information includes at least one implementation step corresponding to the implementation of the interaction purpose and the implementation duration of each implementation step.

[0126] For each interaction record, judging whether the implementation step information meets the preset step optimal condition can include the following operations:

[0127] For each interaction record, the interaction requirement information corresponding to the target user in the interaction record is obtained. The interaction requirement information includes interaction step requirement and interaction duration requirement.

[0128] For each interaction record, it is judged whether the number of implementation steps meets the interaction step requirement and whether the implementation duration of each implementation step meets the interaction duration requirement. When the number of steps meets the interaction step requirement and the implementation duration of each implementation step meets the interaction duration requirement, the key information corresponding to the interaction purpose of the target user is analyzed.

[0129] For each interaction record, the first step number of the interaction step in which the key information is obtained and the second step number of the interaction step in which the interaction purpose is implemented are determined, and it is judged whether the number difference between the first step number and the second step number is greater than a preset number difference threshold. When the number difference is greater than the number difference threshold, it is determined that the implementation step information does not meet the preset step optimal condition.

[0130] In this optional embodiment, optionally, the implementation step information can include at least one implementation step corresponding to the implementation of the interaction purpose and the implementation duration of each implementation step. For each interaction record, the interaction requirement information corresponding to the target user in the interaction record is obtained. The interaction requirement information can include interaction step requirement and interaction duration requirement. The step optimal condition can include the interaction requirement information. The interaction requirement information corresponding to the target user can be used to reflect the interaction experience of the user. That is, when the interaction requirement information is met, it can be determined that the interaction experience of the user is met, and when the requirement information is not met, it can be determined that the interaction experience of the user is not met. Specifically, it can be judged whether the number of implementation steps meets the interaction step requirement and whether the implementation duration of each implementation step meets the interaction duration requirement. When the number of steps does not meet the interaction step requirement or the implementation duration of at least one implementation step does not meet the interaction duration requirement, it can be determined that the interaction experience of the user is not met. When the number of steps meets the interaction step requirement and the implementation duration of each implementation step meets the interaction duration requirement, it can be determined that the interaction experience of the user is met. At this time, the key information corresponding to the interaction purpose of the target user can be analyzed. The key information can include key content used to implement the interaction purpose, such as one or more combinations of keywords, key sentences, key actions, key voices, key images, and key videos. This embodiment is not limited.

[0131] In the optional embodiment, optionally, for each interaction record, a first step number of an interaction step in which the key information is obtained and a second step number of an interaction step in which the interaction purpose is achieved are determined, for example, the first step number of the interaction step in which the key information is obtained can be the second step, the second step number of the interaction step in which the interaction purpose is achieved can be the fifth step, that is, after the key information is obtained in the second step, the third step, the fourth step and the fifth step are passed through, and the interaction purpose corresponding to the key information is achieved, whether the number difference between the first step number and the second step number is greater than a preset number difference threshold value is determined, the number difference in the above example is 5-2=3, when the number difference is greater than the number difference threshold value, it is determined that the implementation step information does not satisfy the preset step optimal condition, the preset number difference threshold value can be set by the user, or can be obtained based on the user's historical use record through big data analysis, for example, the number difference threshold value can be 2, and the embodiment is not limited.

[0132] It can be seen that by implementing the optional embodiment, for each interaction record, the interaction requirement information corresponding to the target user in the interaction record can be obtained, for each interaction record, whether the number of implementation steps satisfies the interaction step requirement and whether the implementation time length of each implementation step satisfies the interaction time length requirement is determined, when the number of steps satisfies the interaction step requirement and the implementation time length of each implementation step satisfies the interaction time length requirement, the key information corresponding to the interaction purpose of the target user is analyzed, the first step number of the interaction step in which the key information is obtained and the second step number of the interaction step in which the interaction purpose is achieved are determined, and whether the number difference between the first step number and the second step number is greater than a preset number difference threshold value is determined, when the number difference is greater than the number difference threshold value, it is determined that the implementation step information does not satisfy the preset step optimal condition, the accuracy and reliability of determining whether the implementation step information satisfies the preset step optimal condition can be improved based on the interaction step and the time length, thereby improving the accuracy and reliability of determining the interaction control strategy to be improved, and the pertinence and accuracy of optimizing the strategy are improved.

[0133] In another optional embodiment, the interaction optimization direction includes a first interaction optimization direction, and further includes a second interaction optimization direction or a third interaction optimization direction.

[0134] Determining the interaction optimization direction corresponding to the interaction control strategy to be improved can include the following operations:

[0135] According to the interaction control strategy to be improved, at least one type of interaction user and the interaction habit corresponding to each type of interaction user are determined;

[0136] For each type of interactive user, according to the interactive step information of the interactive control strategy to be improved, determine the matching degree between each interactive step and the interactive habit of the type of interactive user, and determine whether each matching degree is greater than a preset matching degree threshold;

[0137] Determine the interactive step with a matching degree less than the matching degree threshold as a target interactive step, and determine a first interactive optimization direction corresponding to the interactive control strategy to be improved according to each target interactive step, the first interactive optimization direction including a first interactive step optimization direction;

[0138] When the number of steps of the interactive control strategy to be improved does not meet the interactive step requirement, or the implementation time length of at least one implementation step does not meet the interactive time length requirement, determine a second interactive optimization direction corresponding to the interactive control strategy to be improved, the second interactive optimization direction including a second interactive step optimization direction and / or an interactive time length optimization direction;

[0139] When the number difference of the interactive control strategy to be improved is greater than the number difference threshold, determine a third interactive optimization direction corresponding to the interactive control strategy to be improved according to the number difference, the third interactive optimization direction including a third interactive step optimization direction.

[0140] In this optional embodiment, optionally, the interactive optimization direction can include the first interactive optimization direction, and further include the second interactive optimization direction or the third interactive optimization direction, at least one type of interactive user and the interactive habit corresponding to each type of interactive user can be determined according to the interactive control strategy to be improved, the interactive users can be classified based on information such as age, gender, occupation, and problem type in historical use records, the interactive habit corresponding to each type of interactive user can include one or more combinations of interactive step number habit, interactive time length habit, interactive problem inquiry habit, and emotion change habit in the interactive process, and the present embodiment is not limited.

[0141] In the optional embodiment, optionally, for each type of interactive user, the matching degree between each interactive step and the interactive habit of the type of interactive user can be determined according to the interactive step information of the interactive control strategy to be improved. Specifically, the matching degree between the specific content, step sequence, step duration and the like of the interactive step and the interactive habit of the type of interactive user can be determined, and it is determined whether each matching degree is greater than a preset matching degree threshold. When the matching degree is greater than or equal to the matching degree threshold, it indicates that the step is set reasonably, and this time does not need to be modified. When the matching degree is less than the matching degree threshold, it can be determined that the interactive step with the matching degree less than the matching degree threshold is a target interactive step, that is, the interactive step needs to be optimized. The first interactive optimization direction corresponding to the interactive control strategy to be improved can be determined according to each target interactive step. The first interactive optimization direction includes a first interactive step optimization direction. The first interactive step optimization direction can include an optimization direction such as adjusting the step sequence of the interactive step, modifying the step content, and deleting the interactive step. This embodiment is not limited.

[0142] In the optional embodiment, optionally, when the number of steps of the interactive control strategy to be improved does not meet the interactive step requirement, or the implementation duration of at least one implementation step does not meet the interactive duration requirement, that is, the number of steps of the interactive control strategy to be improved exceeds the step number threshold, or the implementation duration of at least one implementation step exceeds the duration threshold, the second interactive optimization direction corresponding to the interactive control strategy to be improved can be determined. Specifically, the second interactive optimization direction corresponding to the interactive control strategy to be improved can be determined according to the step number, the step number threshold, the implementation duration, and the implementation duration threshold. The second interactive optimization direction can include a second interactive step optimization direction and / or an interactive duration optimization direction. Specifically, the second interactive step optimization direction can include an optimization direction such as adjusting the step sequence of the interactive step, modifying the step content, and deleting the interactive step. The interactive duration optimization direction can include shortening the interactive duration. This embodiment is not limited.

[0143] In the optional embodiment, optionally, when the number difference of the interactive control strategy to be improved is greater than the number difference threshold, it can be indicated that the interactive control strategy meets the interactive experience demand of the user, but there are redundant steps or steps that can be further simplified. At this time, the third interactive optimization direction corresponding to the interactive control strategy to be improved can be determined according to the number difference. The third interactive optimization direction includes a third interactive step optimization direction. The third interactive step optimization direction can include determining a simplifiable step between the first step number of the interactive step of obtaining the key information and the second step number of the interactive step of achieving the interactive purpose, and deleting or simplifying the step. This embodiment is not limited.

[0144] It can be seen that implementing the optional embodiment can determine at least one type of interactive user and the interaction habit corresponding to each type of interactive user according to the interaction control strategy to be improved. For each type of interactive user, the matching degree between each interaction step and the interaction habit of the type of interactive user is determined according to the interaction step information of the interaction control strategy to be improved, and it is determined whether each matching degree is greater than a preset matching degree threshold. The interaction step with a matching degree less than the matching degree threshold is determined as a target interaction step, and the first interaction optimization direction corresponding to the interaction control strategy to be improved is determined according to each target interaction step. When the number of steps of the interaction control strategy to be improved does not meet the interaction step requirement, or the implementation time length of at least one implementation step does not meet the interaction time length requirement, the second interaction optimization direction corresponding to the interaction control strategy to be improved is determined. When the number difference of the interaction control strategy to be improved is greater than the number difference threshold, the third interaction optimization direction corresponding to the interaction control strategy to be improved is determined according to the number difference. Multiple interaction optimization directions can be determined, the accuracy and comprehensiveness of the determined interaction optimization direction are improved, and the interaction process can be simplified to improve the digital human interaction efficiency.

[0145] In yet another optional embodiment, the application of the interaction optimization implementation method for digital humans can further include the following operations:

[0146] determining the deviation information between the interaction result of the interaction control strategy to be improved and the interaction purpose of the target user corresponding to the interaction control strategy to be improved;

[0147] For each interaction step of the interaction control strategy to be improved, it is determined whether the target user of the interaction control strategy to be improved feeds back user negative feedback information in the interaction step to obtain an analysis result. The user negative feedback information includes at least one of negative feedback speech, negative feedback text, negative feedback action, and negative feedback emotion.

[0148] For each interaction step of the interaction control strategy to be improved, when the analysis result indicates that the target user feeds back user negative feedback information in the interaction step, the influence degree of the user negative feedback information on the deviation information is determined.

[0149] And determining the interaction optimization direction corresponding to the interaction control strategy to be improved can include the following operations:

[0150] determining the interaction optimization direction corresponding to the interaction control strategy to be improved according to each influence degree.

[0151] In the optional embodiment, optionally, deviation information between the interaction result of the to-be-improved interaction control strategy and the interaction purpose of the target user corresponding to the to-be-improved interaction control strategy can be determined. The interaction result represents the result finally output by the digital human. The interaction result can deviate from the interaction purpose of the target user. When the deviation information does not exist or the content of the deviation information is less than a preset threshold, it indicates that the interaction result can achieve or meet the interaction purpose of the target user. When the content of the deviation information is greater than the preset threshold, it indicates that the interaction result output by the digital human does not match the interaction purpose of the target user, or the interaction result output by the digital human does not meet or cannot achieve the interaction purpose of the target user. This embodiment is not limited.

[0152] In the optional embodiment, optionally, whether the target user corresponding to the to-be-improved interaction control strategy feeds back user negative feedback information in the interaction step can be determined to obtain an analysis result. The user negative feedback information includes at least one of negative feedback voice, negative feedback text, negative feedback action, and negative feedback emotion. For example, the user interrupts the communication with the digital human through voice, text, action, key control, or changes the content output direction of the digital human. When the analysis result indicates that the target user feeds back the user negative feedback information in the interaction step, the influence degree of the user negative feedback information on the deviation information is determined, that is, the influence degree of the user negative feedback information on the deviation information. Then, the interaction optimization direction corresponding to the to-be-improved interaction control strategy can be determined according to each influence degree. This embodiment is not limited.

[0153] It can be seen that by implementing the optional embodiment, the deviation information between the interaction result of the to-be-improved interaction control strategy and the interaction purpose of the target user corresponding to the to-be-improved interaction control strategy can be determined. For each interaction step of the to-be-improved interaction control strategy, whether the target user corresponding to the to-be-improved interaction control strategy feeds back user negative feedback information in the interaction step can be determined to obtain an analysis result. For each interaction step of the to-be-improved interaction control strategy, when the analysis result indicates that the target user feeds back the user negative feedback information in the interaction step, the influence degree of the user negative feedback information on the deviation information is determined. The accuracy and reliability of determining the response degree can be improved. The interaction optimization direction corresponding to the to-be-improved interaction control strategy can be determined according to each influence degree, the accuracy and comprehensiveness of determining the interaction optimization direction are improved, and then the interaction process can be simplified to improve the digital human interaction efficiency.

[0154] In yet another optional embodiment, the interaction optimization direction includes a fourth interaction optimization direction.

[0155] Determining the interaction optimization direction corresponding to the to-be-improved interaction control strategy can include the following operations:

[0156] In each interaction step of the interaction control strategy to be improved, at least one prediction step of predicting the user's intention by the digital human is screened, and prediction information of each prediction step is determined;

[0157] For each prediction step, user satisfaction information of the target user in at least one interaction step after the prediction step is obtained, and prediction accuracy of the prediction step is determined according to the user satisfaction information and the prediction information of the prediction step;

[0158] According to the prediction accuracy of each prediction step, a fourth interaction optimization direction corresponding to the interaction control strategy to be improved is determined.

[0159] In the optional embodiment, optionally, the interaction optimization direction can include the fourth interaction optimization direction, at least one prediction step of predicting the user's intention by the digital human can be screened in each interaction step of the interaction control strategy to be improved, and prediction information of each prediction step is determined. In the prediction step, the digital human predicts the interaction intention of the user according to the communication content with the target user, and the prediction information can include the predicted content of the digital human, which is not limited in the embodiment.

[0160] In the optional embodiment, optionally, after the digital human outputs the predicted content, user satisfaction information of the target user in at least one interaction step after the prediction step can be obtained. Specifically, feedback information of the user in one or more steps after the prediction step can be obtained, which can include one or more combinations of voice feedback information, text feedback information, action feedback information, and emotional feedback information, and the like. Then, the user satisfaction information of the user is determined. According to the user satisfaction information and the prediction information of the prediction step, the prediction accuracy of the prediction step is determined, and according to the prediction accuracy of each prediction step, a fourth interaction optimization direction corresponding to the interaction control strategy to be improved is determined. The fourth interaction optimization direction can include an interaction optimization direction for the prediction step, which can include adjusting the prediction algorithm, prediction basis in the prediction step, and increasing or reducing the prediction step, which is not limited in the embodiment.

[0161] It can be seen that by implementing the optional embodiment, at least one prediction step of predicting the user's intention by the digital human can be screened in each interaction step of the interaction control strategy to be improved, and prediction information of each prediction step is determined. For each prediction step, user satisfaction information of the target user in at least one interaction step after the prediction step is obtained, and prediction accuracy of the prediction step is determined according to the user satisfaction information and the prediction information of the prediction step. According to the prediction accuracy of each prediction step, a fourth interaction optimization direction corresponding to the interaction control strategy to be improved is determined. The accuracy and comprehensiveness of the determined interaction optimization direction can be improved, and the interaction process can be simplified to improve the interaction efficiency of the digital human.

[0162] Embodiment three

[0163] Please refer to Figure 3 , Figure 3 is a structural schematic diagram of an interactive optimization implementation device for digital people according to an embodiment of the present application. Among them, Figure 3 The interactive optimization implementation device for digital people described can include a smart device for interacting with a user, which can include any of a user device, a smart server, or a smart platform. The user device can have a digital person program built-in for interacting with the user. Specifically, the user device can include, but is not limited to, at least one of a smart phone (Android phone, iOS phone, etc.) with a digital person interaction function, a smart phone number watch, and a smart home device, etc. The smart server can include a local server or a cloud server, and the present application does not limit the embodiment. As Figure 3 shown, the interactive optimization implementation device for digital people can include:

[0164] The collection module 301 is configured to collect interaction records corresponding to a plurality of digital people in a target time period.

[0165] The analysis module 302 is configured to analyze each interaction record to obtain an interaction control strategy to be improved.

[0166] The first determination module 303 is configured to determine an interaction optimization direction corresponding to the interaction control strategy to be improved.

[0167] The optimization module 304 is configured to perform an optimization operation on the interaction control strategy to be improved according to the interaction optimization direction, to update the interaction control strategy to be improved.

[0168] As can be seen, the embodiment Figure 3 The interactive optimization implementation device for digital people described can collect interaction records corresponding to a plurality of digital people in a target time period, analyze each interaction record to obtain an interaction control strategy to be improved, determine an interaction optimization direction corresponding to the interaction control strategy to be improved, and perform an optimization operation on the interaction control strategy to be improved according to the interaction optimization direction, to update the interaction control strategy to be improved. This can improve the user's interaction experience, improve the quality of digital person interaction, improve the digital person service experience and effect, and simplify the interaction process to improve the efficiency of digital person interaction.

[0169] In an optional embodiment, as Figure 4 shown, the analysis module 302 analyzes each interaction record to obtain the specific manner of the interaction control strategy to be improved, which includes:

[0170] For each interaction record, determine the interaction purpose of the target user and the interaction control strategy in the interaction record.

[0171] For each interaction record, the interaction control strategy of the interaction record is analyzed, interaction step information corresponding to the interaction control strategy is determined, and the interaction step information includes at least one interaction step and an interaction time length corresponding to each interaction step;

[0172] For each interaction record, the interaction step information of the interaction record and the interaction purpose of the target user are determined to determine the implementation step information corresponding to the implementation of the interaction purpose in the interaction record;

[0173] For each interaction record, it is determined whether the implementation step information meets the preset step optimal condition, and when the implementation step information does not meet the preset step optimal condition, the interaction control strategy of the interaction record is determined as the interaction control strategy to be improved.

[0174] It can be seen that the implementation Figure 4 The described interaction optimization implementation device applied to digital people can collect interaction records corresponding to a plurality of digital people in a target time period, for each interaction record, determine the interaction purpose of the target user and the interaction control strategy in the interaction record, analyze the interaction control strategy of the interaction record, determine the interaction step information corresponding to the interaction control strategy, determine the implementation step information corresponding to the implementation of the interaction purpose in the interaction record according to the interaction step information of the interaction record and the interaction purpose of the target user, and determine whether the implementation step information meets the preset step optimal condition. When the implementation step information does not meet the preset step optimal condition, the interaction control strategy of the interaction record is determined as the interaction control strategy to be improved, which can improve the accuracy and reliability of determining the interaction control strategy to be improved, and then improve the pertinence and accuracy of optimizing the strategy, determine the interaction optimization direction corresponding to the interaction control strategy to be improved, and perform optimization operation on the interaction control strategy to be improved according to the interaction optimization direction. To update the interaction control strategy to be improved, it can improve the interaction experience of users, improve the quality of digital human interaction, improve the service experience and effect of digital human, and can simplify the interaction process to improve the efficiency of digital human interaction.

[0175] In another optional embodiment, as Figure 4 The implementation step information includes at least one implementation step corresponding to the implementation of the interaction purpose and an implementation time length of each implementation step;

[0176] The specific manner in which the analysis module 302 determines, for each interaction record, whether the implementation step information meets the preset step optimal condition includes:

[0177] For each interaction record, the interaction requirement information corresponding to the target user in the interaction record is obtained, and the interaction requirement information includes interaction step requirements and interaction time length requirements;

[0178] For each interaction record, it is judged whether the number of implementation steps meets the interaction step requirement and whether the implementation time length of each implementation step meets the interaction time length requirement. When the number of steps meets the interaction step requirement and the implementation time length of each implementation step meets the interaction time length requirement, the key information corresponding to the interaction purpose of the target user is analyzed.

[0179] For each interaction record, the first step number of the interaction step in which the key information is obtained and the second step number of the interaction step in which the interaction purpose is implemented are determined, and it is judged whether the number difference between the first step number and the second step number is greater than a preset number difference threshold. When the number difference is greater than the number difference threshold, it is determined that the implementation step information does not meet the preset step optimal condition.

[0180] It can be seen that the implementation Figure 4 The described interaction optimization implementation device for digital people can obtain the interaction requirement information of the target user in each interaction record. For each interaction record, it is judged whether the number of implementation steps meets the interaction step requirement and whether the implementation time length of each implementation step meets the interaction time length requirement. When the number of steps meets the interaction step requirement and the implementation time length of each implementation step meets the interaction time length requirement, the key information corresponding to the interaction purpose of the target user is analyzed. The first step number of the interaction step in which the key information is obtained and the second step number of the interaction step in which the interaction purpose is implemented are determined, and it is judged whether the number difference between the first step number and the second step number is greater than a preset number difference threshold. When the number difference is greater than the number difference threshold, it is determined that the implementation step information does not meet the preset step optimal condition. The accuracy and reliability of judging whether the implementation step information meets the preset step optimal condition can be improved based on the interaction steps and time length, thereby improving the accuracy and reliability of determining the interaction control strategy to be improved, and further improving the pertinence and accuracy of optimizing the strategy.

[0181] In yet another optional embodiment, as Figure 4 shown, the interaction optimization direction includes a first interaction optimization direction, and further includes a second interaction optimization direction or a third interaction optimization direction;

[0182] The specific manner in which the first determination module 303 determines the interaction optimization direction corresponding to the interaction control strategy to be improved includes:

[0183] According to the interaction control strategy to be improved, at least one type of interaction user and the interaction habit corresponding to each type of interaction user are determined;

[0184] For each type of interaction user, according to the interaction step information of the interaction control strategy to be improved, the matching degree between each interaction step and the interaction habit of the type of interaction user is determined, and it is judged whether each matching degree is greater than a preset matching degree threshold.

[0185] Determine an interaction step whose matching degree is less than a matching degree threshold as a target interaction step, and determine a first interaction optimization direction corresponding to the interaction control strategy to be improved based on each target interaction step, where the first interaction optimization direction includes a first interaction step optimization direction;

[0186] When the number of steps of the interaction control strategy to be improved does not meet the interaction step requirement, or the implementation duration of at least one implementation step does not meet the interaction duration requirement, determining a second interaction optimization direction corresponding to the interaction control strategy to be improved, where the second interaction optimization direction includes a second interaction step optimization direction and / or an interaction duration optimization direction;

[0187] When the number difference of the interaction control strategy to be improved is greater than the number difference threshold, a third interaction optimization direction corresponding to the interaction control strategy to be improved is determined according to the number difference, where the third interaction optimization direction includes a third interaction step optimization direction.

[0188] It can be seen that implementation Figure 4 The described interaction optimization implementation device for digital humans can determine at least one type of interactive user and the interaction habits corresponding to each type of interactive user based on the interaction control strategy to be improved. For each type of interactive user, the device determines the degree of match between each interaction step and the interaction habits of that type of interactive user based on the interaction step information of the interaction control strategy to be improved, and determines whether each match is greater than a preset match threshold. Interaction steps with a match less than the match threshold are determined as target interaction steps. Based on each target interaction step, a first interaction optimization direction corresponding to the interaction control strategy to be improved is determined. When the number of steps in the interaction control strategy to be improved does not meet the interaction step requirement, or the implementation duration of at least one implementation step does not meet the interaction duration requirement, a second interaction optimization direction corresponding to the interaction control strategy to be improved is determined. When the number difference of the interaction control strategy to be improved is greater than the number difference threshold, a third interaction optimization direction corresponding to the interaction control strategy to be improved is determined based on the number difference. Multiple interaction optimization directions can be determined, improving the accuracy and comprehensiveness of the determined interaction optimization directions, thereby simplifying the interaction process and improving the interaction efficiency of digital humans.

[0189] In another optional embodiment, Figure 4 As shown, the interactive optimization implementation device applied to digital humans may also include:

[0190] The second determining module 305 is used to determine the deviation information between the interaction result of the interaction control strategy to be improved and the interaction purpose of the target user corresponding to the interaction control strategy to be improved;

[0191] The second determining module 305 is further configured to determine, for each interaction step of the interaction control strategy to be improved, whether the target user of the interaction control strategy to be improved has provided negative user feedback information in the interaction step, and obtain an analysis result, where the negative user feedback information includes at least one of negative feedback voice, negative feedback text, negative feedback action, and negative feedback emotion;

[0192] The second determining module 305 is further configured to determine, for each interaction step of the interaction control strategy to be improved, the degree of influence of the negative user feedback on the deviation information when the analysis result indicates that the target user provides negative user feedback in the interaction step;

[0193] Furthermore, the specific manner in which the first determining module 303 determines the interaction optimization direction corresponding to the interaction control strategy to be improved includes:

[0194] The interaction optimization direction corresponding to the interaction control strategy to be improved is determined according to each impact degree.

[0195] It can be seen that implementation Figure 4 The described interactive optimization implementation device applied to digital humans can determine the deviation information between the interaction results of the interactive control strategy to be improved and the interaction purpose of the target user corresponding to the interactive control strategy to be improved. For each interaction step of the interactive control strategy to be improved, it is determined whether the target user of the interactive control strategy to be improved feeds back negative user feedback information in the interaction step to obtain an analysis result. For each interaction step of the interactive control strategy to be improved, when the analysis result indicates that the target user feeds back negative user feedback information in the interaction step, it is determined the degree of influence of the negative user feedback information on the deviation information, which can improve the accuracy and reliability of determining the response degree, and can determine the interactive optimization direction corresponding to the interactive control strategy to be improved according to each influence degree, thereby improving the accuracy and comprehensiveness of determining the interactive optimization direction, and thus simplifying the interaction process to improve the interaction efficiency of digital humans.

[0196] In another optional embodiment, Figure 4 As shown, the interactive optimization direction includes a fourth interactive optimization direction;

[0197] The specific manner in which the first determining module 303 determines the interaction optimization direction corresponding to the interaction control strategy to be improved includes:

[0198] In each interaction step of the interactive control strategy to be improved, screening at least one prediction step in which the digital human predicts the user's intention, and determining prediction information for each prediction step;

[0199] For each prediction step, obtaining user satisfaction information of the target user in at least one interaction step after the prediction step, and determining the prediction accuracy of the prediction step based on the user satisfaction information and the prediction information of the prediction step;

[0200] According to the prediction accuracy of each prediction step, a fourth interactive optimization direction corresponding to the interactive control strategy to be improved is determined.

[0201] It can be seen that implementation Figure 4 The described interactive optimization implementation device applied to digital humans can screen at least one prediction step in which the digital human predicts the user's intention in each interactive step of the interactive control strategy to be improved, and determine the prediction information of each prediction step. For each prediction step, the device obtains the user satisfaction information of the target user in at least one interactive step after the prediction step, and determines the prediction accuracy of the prediction step based on the user satisfaction information and the prediction information of the prediction step. Based on the prediction accuracy of each prediction step, the fourth interactive optimization direction corresponding to the interactive control strategy to be improved is determined, which can improve the accuracy and comprehensiveness of the determined interactive optimization direction, and further simplify the interactive process to improve the interaction efficiency of the digital human.

[0202] In another optional embodiment, Figure 4 As shown, the optimization module 304 performs optimization operations on the interactive control strategy to be improved according to the interactive optimization direction, and the specific manner of updating the interactive control strategy to be improved includes:

[0203] performing an optimization operation on the interactive control strategy to be improved according to the interactive optimization direction to update the interactive control strategy to be improved;

[0204] Determine the current network status, which includes an online network status or an offline network status;

[0205] When the current network state includes an online network state, the updated interactive control strategy to be improved is updated to each digital human terminal through the network; when the current network state includes an offline network state, the updated interactive control strategy to be improved is stored in the local digital human database.

[0206] It can be seen that implementation Figure 4The described interactive optimization implementation device applied to digital humans can perform optimization operations on the interactive control strategy to be improved according to the interactive optimization direction to update the interactive control strategy to be improved, determine the current network status, and when the current network status includes an online network status, update the updated interactive control strategy to be improved to each digital human terminal through the network. When the current network status includes an offline network status, store the updated interactive control strategy to be improved in the local digital human database. It can process the updated strategy based on the network status, improve the execution efficiency of the optimized strategy, and thus improve the user's interactive experience, enhance the interactive quality of digital humans, and enhance the digital human service experience and effect.

[0207] Example 4

[0208] See also Figure 5 , Figure 5 This is a structural diagram of another device for realizing interactive optimization of digital human disclosed in an embodiment of the present invention. Figure 5 As shown, the interactive optimization implementation device applied to digital humans may include:

[0209] A memory 401 storing executable program code;

[0210] a processor 402 coupled to the memory 401;

[0211] The processor 402 calls the executable program code stored in the memory 401 to execute the steps of the interactive optimization implementation method applied to the digital human described in the first embodiment of the present invention or the second embodiment of the present invention.

[0212] Example 5

[0213] An embodiment of the present invention discloses a computer storage medium storing computer instructions. When the computer instructions are called, they are used to execute the steps of the interactive optimization implementation method applied to digital humans described in the first embodiment or the second embodiment of the present invention.

[0214] Example 6

[0215] An embodiment of the present invention discloses a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program, and the computer program is operable to cause a computer to execute the steps of the interactive optimization implementation method applied to a digital human described in Example 1 or Example 2.

[0216] The apparatus embodiments described above are only illustrative, wherein the modules described as separate components can or can not be physically separated, and the components displayed as modules can or can not be physical modules, i.e., can be located in one place or distributed to multiple network modules. Part or all of the modules can be selected to achieve the purposes of the embodiments according to actual needs. Those skilled in the art can understand and implement without creative labor.

[0217] Through the specific description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be realized by means of software and necessary general hardware platform, and of course can also be realized by hardware. Based on such understanding, the above technical solutions can be embodied in the form of software products, and the computer software products can be stored in a computer readable storage medium, including Read-Only Memory (ROM), Random Access Memory (RAM), Programmable Read-Only Memory (PROM), Erasable Programmable Read-Only Memory (EPROM), One-time Programmable Read-Only Memory (OTPROM), Electrically-Erasable Programmable Read-Only Memory (EEPROM), Compact Disc Read-Only Memory (CD-ROM) or other optical disk storage, magnetic disk storage, magnetic tape storage, or any other computer readable medium capable of carrying or storing data.

[0218] Finally, it should be noted that: the application embodiment discloses only a kind of interactive optimization implementation method and device applied to digital person disclosed by the preferred embodiment of the application, only for describing the technical scheme of the application, not for its limitation;Although the application is described in detail with reference to the foregoing embodiments, those skilled in the art should understand that;It can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement to part of the technical features;And these modifications or replacements do not make the essence of the corresponding technical scheme deviate from the spirit and scope of the technical solutions of the embodiments of the application.

Claims

1. A method for implementing interaction optimization applied to a digital human, characterized in that, The method comprises: collecting interaction records corresponding to a plurality of digital humans in a target time period, and analyzing each of the interaction records to obtain an interaction control strategy to be improved; determining an interaction optimization direction corresponding to the interaction control strategy to be improved, and performing an optimization operation on the interaction control strategy to be improved according to the interaction optimization direction to update the interaction control strategy to be improved; and the analysis of each of the interaction records to obtain the interaction control strategy to be improved comprises: for each of the interaction records, determining an interaction purpose of a target user in the interaction record and an interaction control strategy; for each of the interaction records, analyzing the interaction control strategy of the interaction record to determine interaction step information corresponding to the interaction control strategy, the interaction step information comprising at least one interaction step and an interaction time length corresponding to each of the interaction steps; for each of the interaction records, determining implementation step information corresponding to the interaction purpose in the interaction record according to the interaction step information of the interaction record and the interaction purpose of the target user; for each of the interaction records, determining whether the implementation step information satisfies a preset step optimal condition, and when the implementation step information does not satisfy the preset step optimal condition, determining that the interaction control strategy of the interaction record is the interaction control strategy to be improved; and the implementation step information comprises at least one implementation step corresponding to the interaction purpose and an implementation time length of each of the implementation steps; the determination of whether the implementation step information satisfies the preset step optimal condition for each of the interaction records comprises: for each of the interaction records, obtaining interaction requirement information corresponding to the target user in the interaction record, the interaction requirement information comprising an interaction step requirement and an interaction time length requirement; for each of the interaction records, determining whether the number of steps of the implementation steps satisfies the interaction step requirement and whether the implementation time length of each of the implementation steps satisfies the interaction time length requirement, analyzing key information corresponding to the interaction purpose of the target user when the number of steps satisfies the interaction step requirement and the implementation time length of each of the implementation steps satisfies the interaction time length requirement; for each of the interaction records, determining a first step number of an interaction step in which the key information is obtained and a second step number of an interaction step in which the interaction purpose is implemented, and determining whether a number difference between the first step number and the second step number is greater than a preset number difference threshold value, and when the number difference is greater than the number difference threshold value, determining that the implementation step information does not satisfy the preset step optimal condition.

2. The method for implementing interaction optimization applied to digital people according to claim 1, wherein, The interaction optimization direction comprises a first interaction optimization direction, and further comprises a second interaction optimization direction or a third interaction optimization direction; the determination of the interaction optimization direction corresponding to the interaction control strategy to be improved comprises: determining at least one type of interaction user and an interaction habit corresponding to each type of the interaction user according to the interaction control strategy to be improved; For each type of said interactive user, according to the interaction step information of said to-be-improved interaction control strategy, determine the matching degree between each said interaction step and the interaction habit of this type of interactive user, and determine whether each said matching degree is greater than a preset matching degree threshold; Determine the interaction step whose said matching degree is less than said matching degree threshold as a target interaction step, and according to each said target interaction step, determine the first interaction optimization direction corresponding to said to-be-improved interaction control strategy, said first interaction optimization direction including a first interaction step optimization direction; When the number of steps of said to-be-improved interaction control strategy does not meet said interaction step requirement, or the implementation time length of at least one said implementation step does not meet said interaction time length requirement, determine the second interaction optimization direction corresponding to said to-be-improved interaction control strategy, said second interaction optimization direction including a second interaction step optimization direction and / or an interaction time length optimization direction; When the number difference of said to-be-improved interaction control strategy is greater than said number difference threshold, determine the third interaction optimization direction corresponding to said to-be-improved interaction control strategy according to said number difference, said third interaction optimization direction including a third interaction step optimization direction.

3. The interactive optimization implementation method for digital human according to claim 1 or 2, characterized in that, The method further comprises: Determine the deviation information between the interaction result of said to-be-improved interaction control strategy and the interaction purpose of the target user corresponding to said to-be-improved interaction control strategy; For each said interaction step of said to-be-improved interaction control strategy, determine whether the target user of said to-be-improved interaction control strategy feeds back user negative feedback information in this interaction step to obtain an analysis result, said user negative feedback information including at least one of negative feedback voice, negative feedback text, negative feedback action and negative feedback emotion; For each said interaction step of said to-be-improved interaction control strategy, when said analysis result indicates that said target user feeds back said user negative feedback information in this interaction step, determine the influence degree of said user negative feedback information on said deviation information; And, said determining the interaction optimization direction corresponding to said to-be-improved interaction control strategy comprises: Determine the interaction optimization direction corresponding to said to-be-improved interaction control strategy according to each said influence degree.

4. The interactive optimization implementation method for digital human according to claim 1 or 2, characterized in that, Said interaction optimization direction includes a fourth interaction optimization direction; Said determining the interaction optimization direction corresponding to said to-be-improved interaction control strategy comprises: In each said interaction step of said to-be-improved interaction control strategy, screen at least one prediction step of said digital human predicting user intention, and determine the prediction information of each said prediction step; For each said prediction step, obtain the user satisfaction information of said target user in at least one interaction step after this prediction step, and according to said user satisfaction information and the prediction information of this prediction step, determine the prediction accuracy of this prediction step; According to the prediction accuracy of each said prediction step, determine the fourth interaction optimization direction corresponding to said to-be-improved interaction control strategy.

5. The interactive optimization implementation method for digital human according to claim 1 or 2, characterized in that, The optimization operation is performed on the to-be-improved interaction control strategy according to the interaction optimization direction, so as to update the to-be-improved interaction control strategy, and the method comprises the steps of: The optimization operation is performed on the to-be-improved interaction control strategy according to the interaction optimization direction, so as to update the to-be-improved interaction control strategy; A current network state is determined, and the current network state comprises an online network state or an offline network state; When the current network state comprises the online network state, the updated to-be-improved interaction control strategy is updated to each digital human terminal through a network, and when the current network state comprises the offline network state, the updated to-be-improved interaction control strategy is stored in a local digital human database.

6. An apparatus for implementing interaction optimization applied to a digital human, characterized by, The device comprises: A collection module is configured to collect interaction records corresponding to a plurality of digital humans in a target time period; An analysis module is configured to analyze each interaction record to obtain a to-be-improved interaction control strategy; A first determination module is configured to determine an interaction optimization direction corresponding to the to-be-improved interaction control strategy; An optimization module is configured to perform an optimization operation on the to-be-improved interaction control strategy according to the interaction optimization direction, so as to update the to-be-improved interaction control strategy; The analysis module analyzes each interaction record to obtain a to-be-improved interaction control strategy, and the manner comprises the following steps: For each interaction record, the interaction purpose of a target user in the interaction record and an interaction control strategy are determined; For each interaction record, the interaction control strategy of the interaction record is analyzed to determine interaction step information corresponding to the interaction control strategy, and the interaction step information comprises at least one interaction step and an interaction time length corresponding to each interaction step; For each interaction record, the implementation step information corresponding to the implementation of the interaction purpose in the interaction record is determined according to the interaction step information of the interaction record and the interaction purpose of the target user; For each interaction record, it is determined whether the implementation step information satisfies a preset step optimization condition, and when the implementation step information does not satisfy the preset step optimization condition, the interaction control strategy of the interaction record is determined as a to-be-improved interaction control strategy; The implementation step information comprises at least one implementation step corresponding to the implementation of the interaction purpose and an implementation time length of each implementation step; The analysis module analyzes each interaction record to obtain a to-be-improved interaction control strategy, and the manner comprises the following steps: For each interaction record, interaction requirement information corresponding to the target user in the interaction record is obtained, and the interaction requirement information comprises interaction step requirements and interaction time length requirements; For each interaction record, it is determined whether the number of steps of the implementation step satisfies the interaction step requirements and whether the implementation time length of each implementation step satisfies the interaction time length requirements, and when the number of steps satisfies the interaction step requirements and the implementation time length of each implementation step satisfies the interaction time length requirements, key information corresponding to the interaction purpose of the target user is analyzed; For each of the interaction records, a first step number of an interaction step in which the key information is acquired and a second step number of an interaction step in which the interaction purpose is achieved are determined, and it is judged whether a number difference between the first step number and the second step number is greater than a preset number difference threshold value. When the number difference is greater than the number difference threshold value, it is determined that the implementation step information does not satisfy a preset step optimal condition.

7. An apparatus for implementing interaction optimization applied to a digital human, characterized by, The apparatus comprises: a memory storing executable program codes; a processor coupled with the memory; the processor invokes the executable program codes stored in the memory to execute the method for optimizing implementation of interaction of a digital person according to any one of claims 1-5.

8. A computer storage medium, characterized in that, The computer storage medium stores computer instructions, which when invoked, are used to execute the method for optimizing implementation of interaction of a digital person according to any one of claims 1-5.

Images