A seamless handover and context synchronization system and method in a human-robot collaboration process of a customer service robot

Abstract

The application relates to a seamless switching and context synchronization system and method in a customer service man-machine cooperation process, and deep learning is provided in the application, and more particularly relates to the field of electric digital data processing. The system comprises: a big data acquisition device for acquiring big data track communication content of each past time segment of a set e-commerce application before a current time; and a data analysis device for intelligently analyzing the number of intelligent robots and the number of artificial customer services required for seamless switching in the current time segment based on the big data track communication content by using an intelligent customer service data analysis model. By the application, in the face of the technical problem that seamless switching in the customer service man-machine cooperation process is difficult to achieve in different time segments, a deep learning model can be used to intelligently analyze the number of various customer services required for seamless switching in each time segment according to various types of basic information acquired by big data, so that the above technical problem is solved.

Description

Technical Field

[0001] The deep learning proposed in this invention relates more specifically to the field of electronic digital data processing, and more specifically to a seamless switching and context synchronization system and method for human-machine collaboration in customer service robots. Background Technology

[0002] Deep learning ensures the reliability and stability of the obtained deep learning models, enabling them to be widely applied in various sub-application areas of digital data processing, such as customer service in various e-commerce applications.

[0003] In the past, businesses primarily provided customer service through online consultations and telephone calls. Now, in the mobile internet era, users are more accustomed to communicating with businesses via instant messaging (IM), especially e-commerce companies that use e-commerce applications as their user service interface, expecting ubiquitous customer service. Therefore, as an effective replacement for human customer service, intelligent chatbots, through pre-defined communication frameworks, can significantly reduce the workload of human customer service representatives. The current customer service process design for various e-commerce applications primarily follows the pattern of using an intelligent chatbot first, followed by a human representative.

[0004] For example, Chinese invention patent publication CN 120782456A proposes an AI-powered dynamic monitoring and automatic recovery system for e-commerce intelligent customer service, relating to the field of e-commerce service optimization technology. The system includes: a problem urgency judgment module that calculates problem urgency values ​​based on multi-dimensional evaluation indicators to dynamically assess problem priority; a human customer service switching module that uses a dynamic weight allocation mechanism to assign problems to human agents with corresponding professional skills; a dialogue context analysis module that analyzes the dialogue content between customer service representatives and users in real time to generate a dialogue satisfaction curve; an AI customer service switching module that, when the satisfaction curve stabilizes, transfers the dialogue to AI customer service for continued service; and an optimization module that optimizes service strategies based on dialogue history and a knowledge base. This application constructs an efficient and intelligent service mechanism from aspects such as urgency assessment, collaborative service between human and AI customer service, and dynamic optimization, improving the response efficiency and user experience of the customer service system.

[0005] For example, Chinese invention patent publication CN 115620709A proposes an intelligent customer service system for e-commerce platforms based on big data, specifically relating to the field of e-commerce technology. This application converts and analyzes the text, voice, and image information sent by users, and further performs question prediction processing based on the sent voice, text, and images. Based on similar question matching analysis of Internet big data information, it can obtain more relevant language to meet the needs of users, satisfy the normal interaction needs between users, and reflect the intelligent and autonomous interaction purpose of this system. Moreover, through repeated conversations, human customer service can intervene, and the system can achieve memory learning. It can also autonomously filter key information during shift changes, helping shift changers to understand the problem more intuitively and quickly, obtain timely responses, and save operation time. This enables the system to achieve intelligent human-computer interaction, improve user experience and service quality, and reduce user churn rate.

[0006] Therefore, it is evident that the aforementioned existing technical solutions cannot achieve a seamless switching effect between intelligent robots and human customer service in all different time segments. The difficulty lies in determining the appropriate scale and ratio of intelligent robots and human customer service in different time segments to achieve a seamless switching effect that improves user communication experience while conserving limited customer service resources. Consequently, in existing e-commerce applications, the switching between intelligent robots and human customer service may be disrupted in certain time segments, either because the large number of intelligent robots and human customer service providers is not configured in an appropriate ratio, or because the ratio is appropriate but not sufficient to meet the current user communication needs. Summary of the Invention

[0007] To address the technical problems in existing technologies, this invention provides a seamless switching and context synchronization system and method for human-machine collaboration in customer service robots. This involves designing a customized deep learning model for a specific e-commerce application and collecting data on the total number of users accessing the application across multiple past time segments, the consumption of various customer service resources, user access trajectories, customer service communication content, and other auxiliary information. Based on this collected data, the deep learning model intelligently analyzes the number of intelligent robots and human customer service representatives required for seamless switching in the human-machine collaboration process within the current time segment of the e-commerce application. Based on the intelligent analysis results, the system executes seamless switching and context synchronization for the human-machine collaboration process within the current time segment of the e-commerce application. This improves the customer service quality of the e-commerce application while minimizing the waste of limited intelligent robot and human customer service resources.

[0008] According to one aspect of the present invention, a seamless switching and context synchronization system for human-machine collaboration in a customer service robot is provided, the system comprising:

[0009] Big data collection equipment is used to collect access trajectory information and customer service communication data of each user in each past time segment before the current moment of a set e-commerce application, so as to serve as the big data trajectory communication content of that past time segment.

[0010] A parameter capture device used to capture seamless switching data for each past time segment before the current moment in a specified e-commerce application;

[0011] Information parsing equipment, used to parse multiple e-commerce related information for setting up e-commerce applications;

[0012] A continuous reconfiguration device is used to perform deep learning on a BP neural network to obtain an intelligent customer service data analysis model.

[0013] The data analysis equipment is connected to the big data acquisition equipment, parameter capture equipment, information parsing equipment, and continuous reconstruction equipment, respectively. It is used to use the intelligent customer service data analysis model to analyze multiple e-commerce related information of the set e-commerce application, the duration of time segments, the total number of users accessing the set e-commerce application in multiple past time segments before the current moment, seamless switching data, and big data trajectory communication content to achieve the number of intelligent robots and human customer service representatives required for seamless switching in the human-machine collaboration process of the customer service robot within the current moment segment.

[0014] Dynamically configure devices, connect to data analysis devices, and configure seamless switching and context synchronization strategies for e-commerce applications based on intelligent analysis results at the current moment, segmented into current time periods.

[0015] According to another aspect of the present invention, a method for seamless switching and context synchronization during human-machine collaboration in a customer service robot is provided, the method comprising:

[0016] Collect access trajectory information and customer service communication data for each user in each past time segment before the current moment in the e-commerce application, as well as the big data trajectory communication content for that past time segment.

[0017] Capture seamless switching data for every past time segment before the current moment in the e-commerce application;

[0018] Analyze and configure multiple e-commerce association information settings for e-commerce applications;

[0019] Deep learning is performed on a BP neural network to obtain an intelligent customer service data analysis model;

[0020] The intelligent customer service data analysis model is based on multiple e-commerce related information of the e-commerce application, the duration of time segments, the total number of users accessing the e-commerce application in multiple past time segments before the current moment, seamless switching data, and big data trajectory communication content. It intelligently analyzes the number of intelligent robots and human customer service representatives required to achieve seamless switching in the human-machine collaboration process of the customer service robot within the current time segment.

[0021] Based on intelligent analysis results, the e-commerce application is configured with a seamless switching and context synchronization strategy for the current time segment.

[0022] Therefore, it can be seen that the present invention has at least the following five significant technical advancements:

[0023] First, a deep learning model is used to analyze the number of intelligent robots and human customer service representatives required for seamless switching in the human-machine collaboration process of the e-commerce application within the current time segment. This is based on the total number of users accessing the application, the consumption of various customer service resources, the big data access trajectory of users, the big data customer service communication content, and other auxiliary information.

[0024] Secondly, a customized intelligent customer service data analysis model is designed for e-commerce applications as a deep learning model. This model is used to perform intelligent analysis on the number of intelligent robots and human customer service representatives required for seamless switching in the human-machine collaboration process of the e-commerce application within the current time segment. The intelligent customer service data analysis model is a BP neural network that has undergone multiple learning cycles exceeding a preset number of cycles. The numerical trend of the number of learning cycles is consistent with the numerical trend of the number of registered merchants in the e-commerce application. The BP neural network uses the Sigmoid function as the activation function. This allows for the design of intelligent customer service data analysis models with different customized structures for different e-commerce applications, ensuring the reliability and stability of the intelligent analysis results.

[0025] Furthermore, in each learning iteration of the BP neural network, the known number of intelligent robots and human customer service representatives required for seamless switching in the human-customer collaboration process within a certain time segment of the e-commerce application are used as the two outputs of the BP neural network. Multiple e-commerce related information of the e-commerce application, the duration of the time segment, the total number of users accessing the e-commerce application in multiple past time segments before the stated time segment, seamless switching data, and big data trajectory communication content are used as multiple inputs of the BP neural network to complete this learning iteration, thus ensuring the learning effectiveness of each iteration of the BP neural network.

[0026] Furthermore, to perform intelligent analysis on the number of intelligent robots and human customer service representatives required for seamless switching in the human-machine collaboration process of the e-commerce application within the current time segment, several basic information was introduced. These included setting multiple e-commerce related information of the e-commerce application, the duration of the time segment, the total number of users accessing the e-commerce application in multiple past time segments before the current time, seamless switching data, and big data trajectory communication content. The comprehensive and sufficient selection of the above-mentioned basic information further ensured the reliability and stability of the intelligent analysis results.

[0027] Finally: Specifically, the system collects access trajectory information and customer service communication data for each user within each past time segment before the current moment in the e-commerce application. This data serves as the big data trajectory communication content for each past time segment. The system also collects the access duration, access sequence number, product type number, product return rate, product quantity, and product price for each product visited by each user within each past time segment, and the binary sequence and total number of segments for each user's interactions with the AI ​​chatbot within each past time segment. The system uses a binary sequence of ASCII values ​​representing each character in a segment of interaction with human customer service within a given time period, along with the total number of segments, as customer service interaction data for that user within that time period. Simultaneously, it captures the number of intelligent robots and human customer service representatives used for seamless switching between human and customer service robots in each past time period before the current moment, serving as seamless switching data for that time period. Furthermore, it sets various e-commerce-related information for the application, including the number of registered users, registered merchants, total number of product types sold, the fastest processing speed of the server cluster, and the maximum storage capacity. This allows for the design of a customized data structure to comprehensively and fully select from multiple fundamental information sets. Attached Figure Description

[0028] The embodiments of the present invention will now be described with reference to the accompanying drawings, wherein:

[0029] Figure 1 This is a schematic diagram of a working scenario for a seamless switching and context synchronization system and method for human-machine collaboration in a customer service robot according to the present invention.

[0030] Figure 2 This is an internal structure diagram of a seamless switching and context synchronization system in the human-machine collaboration process of a customer service robot, as shown in the first embodiment of the present invention.

[0031] Figure 3 This is an internal structure diagram of a seamless switching and context synchronization system in the human-machine collaboration process of a customer service robot, as shown in the second embodiment of the present invention.

[0032] Figure 4 This is an internal structure diagram of a seamless switching and context synchronization system in the human-machine collaboration process of a customer service robot, as shown in the third embodiment of the present invention.

[0033] Figure 5This is an internal structure diagram of a seamless switching and context synchronization system in the human-machine collaboration process of a customer service robot, as shown in the fourth embodiment of the present invention.

[0034] Figure 6 This is an internal structure diagram of a seamless switching and context synchronization system in the human-machine collaboration process of a customer service robot, as shown in the fifth embodiment of the present invention.

[0035] Figure 7 This is a flowchart illustrating the steps of a seamless switching and context synchronization method in the human-machine collaboration process of a customer service robot according to the sixth embodiment of the present invention. Detailed Implementation

[0036] like Figure 1 The diagram illustrates a working scenario of a seamless switching and context synchronization system and method for human-machine collaboration in a customer service robot, according to the present invention. The deep learning proposed in this invention specifically relates to the field of electronic digital data processing.

[0037] To achieve the present invention, the following technical processes are specifically proposed:

[0038] Technical Process A: For a given e-commerce application, a customized intelligent customer service data analysis model is designed as a deep learning model. This model is used to perform intelligent analysis on the number of intelligent robots and human customer service representatives required for seamless switching between human and customer service robot collaboration within the current time segment of the e-commerce application. Figure 1 As shown;

[0039] For example, following the customer service communication process of first providing intelligent robot service and then human customer service taking over, a small set time threshold can be set. If the switching time consumed by the intelligent robot and human customer service during a certain time segment is less than or equal to the set time threshold, it is determined that a seamless switching in the human-machine collaboration process of the customer service robot has been achieved within that certain time segment.

[0040] Specifically, the structural customization of the intelligent customer service data analysis model designed for e-commerce applications is mainly reflected in the following aspects:

[0041] First: The intelligent customer service data analysis model is a BP neural network that has undergone multiple learning iterations exceeding a preset number of times;

[0042] Specifically, because the preset number of learning iterations is relatively large, a BP neural network that has completed more than the preset number of learning iterations is a BP neural network that has completed a massive number of learning iterations, that is, a BP neural network that has completed deep learning, and belongs to a deep learning model.

[0043] Second: In the intelligent customer service data analysis model, the numerical trend of the number of learning iterations performed by the BP neural network is consistent with the numerical trend of the number of registered merchants in the set e-commerce application.

[0044] For example, if the number of registered merchants in the e-commerce application is set to 5,000, the number of learning sessions is selected to be 1,000; if the number of registered merchants in the e-commerce application is set to 6,000, the number of learning sessions is selected to be 1,200; if the number of registered merchants in the e-commerce application is set to 7,000, the number of learning sessions is selected to be 1,400; if the number of registered merchants in the e-commerce application is set to 8,000, the number of learning sessions is selected to be 1,600, and so on.

[0045] Third: In the intelligent customer service data analysis model, the BP neural network selects the Sigmoid function as the activation function;

[0046] Fourth: In each learning process of the BP neural network, the number of intelligent robots and human customer service representatives required for seamless switching in the human-machine collaboration process of the e-commerce application within a certain time segment are taken as the two output contents of the BP neural network. Multiple e-commerce related information of the e-commerce application, the duration of the time segment, the total number of users accessing the e-commerce application in multiple past time segments before the certain time segment, seamless switching data, and big data trajectory communication content are taken as the multiple input contents of the BP neural network to complete this learning process, thereby ensuring the learning effect of the BP neural network in each learning process.

[0047] In this way, we can design intelligent customer service data analysis models with different customized structures for different e-commerce applications, ensuring the reliability and stability of intelligent analysis results.

[0048] Technical Process B: To perform intelligent analysis of the number of intelligent robots and human customer service representatives required for seamless switching in the human-machine collaboration process of the e-commerce application within the current time segment, several basic information items were introduced;

[0049] Specifically, such as Figure 1 As shown, the various basic information includes setting multiple e-commerce association information for the e-commerce application, duration of time segments, total number of users accessing the e-commerce application in multiple past time segments before the current moment, seamless switching data, and big data trajectory communication content.

[0050] More specifically, the number of selected past time segments is positively correlated with the total number of product types sold by the e-commerce application. For example, if the total number of product types sold by the e-commerce application is set to 20,000, the number of selected past time segments is 10; if the total number of product types sold by the e-commerce application is set to 30,000, the number of selected past time segments is 15; if the total number of product types sold by the e-commerce application is set to 40,000, the number of selected past time segments is 20; if the total number of product types sold by the e-commerce application is set to 50,000, the number of selected past time segments is 25, and so on.

[0051] More specifically, the system collects access trajectory information and customer service communication data for each user in each past time segment before the current moment of the e-commerce application, as well as customer service communication data for each user, to serve as the big data trajectory communication content for that past time segment.

[0052] In the big data trajectory communication content, the visit duration, visit sequence number, product type number, product return rate, product quantity, and product sales price of each product visited by each user in each past time segment are used as the visit trajectory information of the user in that past time segment. The binary value sequence formed by sequentially connecting the ASCII values ​​of each component character of each paragraph of communication between each user and the intelligent robot in each past time segment, and the total number of paragraphs, and the binary value sequence formed by sequentially connecting the ASCII values ​​of each component character of each paragraph of communication between each user and human customer service in each past time segment, and the total number of paragraphs, are used as the customer service communication data of the user in that past time segment.

[0053] More specifically, the number of intelligent robots and human customer service representatives used to achieve seamless switching in the human-machine collaboration process of the e-commerce application in each past time segment before the current moment is captured as seamless switching data for that past time segment.

[0054] More specifically, setting various e-commerce related information for the e-commerce application includes setting the number of registered users, the number of registered merchants, the total number of product types sold, the fastest processing speed of the server cluster, and the maximum storage capacity.

[0055] In this way, by comprehensively and fully selecting the above-mentioned basic information, the reliability and stability of the intelligent analysis results are further guaranteed;

[0056] Technical Process C: Using the intelligent customer service data analysis model designed with a customized structure for the e-commerce application based on Technical Process A, and based on multiple basic information collected from big data in Technical Process B, the model intelligently analyzes and sets the required number of intelligent robots and human customer service representatives for seamless switching in the human-machine collaboration process within the current time segment of the e-commerce application. Figure 1 As shown;

[0057] In this way, for any time segment of the e-commerce application, the specific quantities of the two types of customer service resources required for seamless switching in the human-machine collaboration process of the customer service robot can be intelligently analyzed and obtained. This allows the system to directly provide the scale of the two types of customer service resources and their relative proportions, thereby meeting the requirements for seamless switching in the human-machine collaboration process of the customer service robot. This avoids situations where seamless switching is difficult to achieve due to only providing a rough proportion or only providing two large-scale customer service resource quantities.

[0058] Technical Process D: Based on the intelligent analysis results of Technical Process C, configure the corresponding seamless switching and context synchronization strategies for the e-commerce application segmented at the current moment;

[0059] Specifically, the e-commerce application is configured with the number of intelligent robots and human customer service representatives obtained from intelligent analysis for the current time segment, and the context synchronization of the communication segments when switching between intelligent robots and human customer service representatives is configured.

[0060] More specifically, configuring context synchronization of communication segments when switching between the intelligent chatbot and human customer service includes: when switching between the intelligent chatbot and human customer service, resending the most recently sent communication segment by the user to the customer service representative being switched.

[0061] Therefore, through the coordinated operation of the above-mentioned technical processes, this invention can use a deep learning model to perform intelligent analysis on the number of intelligent robots and human customer service representatives required for seamless switching in the human-machine collaboration process of the customer service robot within the current time segment. This analysis is based on the total number of users accessing the e-commerce application in multiple past time segments before the current time, the consumption of various customer service resources, the big data access trajectory of users, the big data customer service communication content, and other auxiliary information. Based on the intelligent analysis results, the invention performs seamless switching and context synchronization in the human-machine collaboration process of the customer service robot within the current time segment of the e-commerce application. This improves the customer service quality of the e-commerce application while minimizing the waste of limited intelligent robot customer service resources and human customer service resources.

[0062] The key points of this invention are: synchronous intelligent analysis of the number of intelligent robots and human customer service representatives required for seamless switching to avoid situations where seamless switching is difficult to achieve due to providing only a rough ratio or only two large-scale customer service resource quantities; customized structural design of different intelligent customer service data analysis models for different e-commerce applications; big data collection of multiple basic information for intelligent analysis; customized design of data structures for multiple basic information; and targeted selection of the number of past time segments that are positively correlated with the total number of sales product types in the set e-commerce application.

[0063] The following will describe in detail, by way of embodiments, a seamless switching and context synchronization system and method for human-machine collaboration in a customer service robot according to the present invention.

[0064] First Embodiment

[0065] Figure 2 This is an internal structure diagram of a seamless switching and context synchronization system in the human-machine collaboration process of a customer service robot, as shown in the first embodiment of the present invention.

[0066] like Figure 2 As shown, the seamless switching and context synchronization system for the human-machine collaboration process of the customer service robot includes the following components:

[0067] Big data collection equipment is used to collect access trajectory information and customer service communication data of each user in each past time segment before the current moment of a set e-commerce application, so as to serve as the big data trajectory communication content of that past time segment.

[0068] Here, through big data collection equipment, the access trajectory and customer service communication of users in each past time segment of the set e-commerce application are collected. The collected results are used for the synchronous intelligent analysis of the number of two types of customer service resources to achieve seamless switching in the current time segment.

[0069] Conceptually, the seamless switching between intelligent robots and human customer service is a relative concept, referring to a very short switching time. For example, if the switching time consumed by the intelligent robot and human customer service during a certain time segment is less than or equal to a set time threshold, it is determined that a seamless switching in the human-machine collaboration process of the customer service robot has been achieved within that time segment. Here, the value of the set time threshold is very small.

[0070] Accordingly, the concept of context synchronization in this invention refers to the process of resending the latest communication segment sent by the user to the switched customer service provider (usually a human customer service representative) when switching between the intelligent robot and the human customer service representative.

[0071] A parameter capture device used to capture seamless switching data for each past time segment before the current moment in a specified e-commerce application;

[0072] For example, if the current time is 5:00 PM, the current time interval is from 5:00 PM to 5:05 PM, and the multiple past time segments before the current time are multiple past time segments before 5:00 PM, such as 4:55 PM to 5:00 PM, 4:50 PM to 4:55 PM, 4:45 PM to 4:50 PM, 4:40 PM to 4:45 PM, 4:35 PM to 4:40 PM, 4:30 PM to 4:35 PM, 4:25 PM to 4:30 PM, 4:20 PM to 4:25 PM, 4:15 PM to 4:20 PM, and 4:10 PM to 4:15 PM, a total of 10 past time segments;

[0073] Information parsing equipment, used to parse multiple e-commerce related information for setting up e-commerce applications;

[0074] Specifically, parsing and setting multiple e-commerce related information of the e-commerce application includes: selecting multiple information collection components to collect multiple e-commerce related information of the e-commerce application respectively. The multiple e-commerce related information of the e-commerce application, like other basic information used for intelligent analysis, can be regarded as a kind of big data to be collected.

[0075] A continuous reconfiguration device is used to perform deep learning on a BP neural network to obtain an intelligent customer service data analysis model.

[0076] For example, numerical simulation mode can be used to test and simulate the model building process of performing deep learning on a BP neural network to obtain an intelligent customer service data analysis model;

[0077] The data analysis equipment is connected to the big data acquisition equipment, parameter capture equipment, information parsing equipment, and continuous reconstruction equipment, respectively. It is used to use the intelligent customer service data analysis model to analyze multiple e-commerce related information of the set e-commerce application, the duration of time segments, the total number of users accessing the set e-commerce application in multiple past time segments before the current moment, seamless switching data, and big data trajectory communication content to achieve the number of intelligent robots and human customer service representatives required for seamless switching in the human-machine collaboration process of the customer service robot within the current moment segment.

[0078] Dynamically configure devices, connect to data analysis devices, and configure seamless switching and context synchronization strategies for e-commerce applications based on intelligent analysis results at the current moment, segmented into current time periods.

[0079] Among them, the seamless switching and context synchronization strategy for the e-commerce application segmented based on the intelligent analysis results at the current moment includes: configuring the number of intelligent robots and human customer service representatives obtained by intelligent analysis for the e-commerce application segmented based on the current moment at the current moment, and configuring the context synchronization of the communication segments of intelligent robots and human customer service representatives when switching.

[0080] Among them, capturing seamless switching data of the e-commerce application in each past time segment before the current time includes: capturing the number of intelligent robots and human customer service representatives used to achieve seamless switching in the human-machine collaboration process of the customer service robot in each past time segment before the current time, as the seamless switching data of that past time segment.

[0081] Specifically, for a set e-commerce application, if the switching time consumed when switching between the intelligent robot and the human customer service is less than or equal to the set time threshold within a certain time segment, it is determined that a seamless switching in the human-machine collaboration process of the customer service robot has been achieved within that certain time segment.

[0082] For example, for a set e-commerce application, when the switching time consumed by the intelligent robot and human customer service during a certain time segment is less than or equal to a set time threshold, it is determined that the seamless switching in the human-machine collaboration process of the customer service robot is achieved within a certain time segment, including: the set time threshold is small enough, for example, at the millisecond level;

[0083] The configuration of context synchronization of communication segments when switching between intelligent chatbots and human customer service includes: when switching between intelligent chatbots and human customer service, the latest communication segment sent by the user will be resent to the customer service representative being switched.

[0084] Specifically, when switching between the intelligent robot and the human customer service, the most recently sent communication paragraphs by the user will be resent to the customer service representative being switched. This can automatically resend the preceding text of the user-customer service communication content, improve the speed and efficiency of the switched customer service representative getting into working mode, and also reduce the burden of resending for the user.

[0085] Among them, setting multiple e-commerce related information for the e-commerce application includes setting the number of registered users, the number of registered merchants, the total number of product types sold, the fastest computing speed of the server cluster, and the maximum storage capacity.

[0086] For example, the fastest computing speed using a server cluster can be expressed in terms of operations per second, and the maximum storage capacity using a server cluster can be expressed in terms of MB.

[0087] The current time segment starts at the current moment, and the current time segment and multiple past time segments before the current moment form a complete time interval on the time axis, and the duration of each time segment is equal.

[0088] For example, if the current time is 5:00 PM, and the current time interval is from 5:00 PM to 5:05 PM, then the previous time segments before the current time are the 10 previous time segments before 5:00 PM. The current time segment and the previous time segments before the current time together form a complete time interval from 4:10 PM to 5:05 PM on the timeline.

[0089] Among them, performing deep learning on the BP neural network to obtain an intelligent customer service data analysis model includes: performing multiple learning operations on the BP neural network more than a preset number of times, so as to obtain the BP neural network after multiple learning operations and use it as the output of the intelligent customer service data analysis model, and the numerical trend of the number of learning operations is consistent with the numerical trend of the number of registered merchants in the set e-commerce application.

[0090] For example, the trend of the number of times the learning is performed is consistent with the trend of the number of registered merchants in the e-commerce application, including: setting the number of registered merchants in the e-commerce application to 5,000 and selecting 1,000 times of learning; setting the number of registered merchants in the e-commerce application to 6,000 and selecting 1,200 times of learning; setting the number of registered merchants in the e-commerce application to 7,000 and selecting 1,400 times of learning; setting the number of registered merchants in the e-commerce application to 8,000 and selecting 1,600 times of learning, and so on.

[0091] In each learning iteration of the BP neural network, the number of intelligent robots and human customer service representatives required for seamless switching in the human-machine collaboration process of the e-commerce application within a certain time segment are taken as the two outputs of the BP neural network. The multiple e-commerce related information of the e-commerce application, the duration of the time segment, the total number of users accessing the e-commerce application in multiple past time segments before the certain time segment, seamless switching data, and big data trajectory communication content are taken as the multiple inputs of the BP neural network to complete this learning iteration of the BP neural network.

[0092] Specifically, the visit duration, visit sequence number, product type number, product return rate, product quantity, and product sales price of each user within each past time segment are used as the visit trajectory information for that user within that past time segment.

[0093] Specifically, the access sequence number of the first product a user visits in a certain past time segment is 1, the access sequence number of the second product a user visits in a certain past time segment is 2, and so on.

[0094] In addition, the binary numerical sequence and the total number of segments formed by sequentially connecting the ASCII values ​​corresponding to each component character of each segment of the interaction between each user and the intelligent robot in each past time segment, and the binary numerical sequence and the total number of segments formed by sequentially connecting the ASCII values ​​corresponding to each component character of each segment of the interaction between each user and the human customer service representative in each past time segment, are used as the customer service interaction data corresponding to the user in that past time segment.

[0095] Second Embodiment

[0096] Figure 3 This is an internal structure diagram of a seamless switching and context synchronization system in the human-machine collaboration process of a customer service robot, as shown in the second embodiment of the present invention.

[0097] like Figure 3 As shown, compared to Figure 2 The seamless switching and context synchronization system for the human-machine collaboration process of the customer service robot also includes:

[0098] A wireless transmission device, connected to a dynamic configuration device, is used to receive seamless handover and context synchronization policies for the current time segment of the set e-commerce application, and transmits the seamless handover and context synchronization policies for the current time segment of the set e-commerce application to the unified control center corresponding to the set e-commerce application at a remote location via a wireless communication link.

[0099] Specifically, the seamless switching and context synchronization strategy for the current segmented e-commerce application will be transmitted to the unified control center corresponding to the remote e-commerce application via a wireless communication link. This includes using a wireless communication link based on frequency division duplex communication mode or time division duplex communication mode.

[0100] The seamless switching and context synchronization strategy for the current segmented e-commerce application will be transmitted via wireless communication link to the unified control center corresponding to the remote e-commerce application. This includes using various big data service network elements as computing nodes within the unified control center corresponding to the remote e-commerce application.

[0101] Third Embodiment

[0102] Figure 4This is an internal structure diagram of a seamless switching and context synchronization system in the human-machine collaboration process of a customer service robot, as shown in the third embodiment of the present invention.

[0103] like Figure 4 As shown, compared to Figure 2 The seamless switching and context synchronization system for the human-machine collaboration process of the customer service robot also includes:

[0104] The real-time display device is set in the user client device running the set e-commerce application and connected to the data analysis device. It is used to receive the number of intelligent robots and human customer service representatives required for seamless switching in the human-machine collaboration process of the customer service robot within the current time segment.

[0105] The real-time display device is also used to simultaneously display the number of intelligent robots and human customer service representatives received;

[0106] For example, a touch screen integrated into the user client device can be selected to receive the number of intelligent robots and human customer service representatives required for seamless switching in the human-machine collaboration process of the customer service robot within the current time segment, and simultaneously display the received number of intelligent robots and human customer service representatives.

[0107] Fourth embodiment

[0108] Figure 5 This is an internal structure diagram of a seamless switching and context synchronization system in the human-machine collaboration process of a customer service robot, as shown in the fourth embodiment of the present invention.

[0109] like Figure 5 As shown, compared to Figure 2 The seamless switching and context synchronization system for the human-machine collaboration process of the customer service robot also includes:

[0110] A model storage device, connected to a continuous reconstruction device, is used to receive intelligent customer service data analysis models and store the intelligent customer service data analysis models.

[0111] For example, an MMC storage chip or a TF storage chip can be selected to implement the model storage device, which is used to receive the intelligent customer service data analysis model and realize the model storage of the intelligent customer service data analysis model;

[0112] The process of receiving and storing intelligent customer service data analysis models includes: storing the intelligent customer service data analysis models by storing various model parameters of the intelligent customer service data analysis models.

[0113] Fifth embodiment

[0114] Figure 6This is an internal structure diagram of a seamless switching and context synchronization system in the human-machine collaboration process of a customer service robot, as shown in the fifth embodiment of the present invention.

[0115] like Figure 6 As shown, compared to Figure 2 The seamless switching and context synchronization system for the human-machine collaboration process of the customer service robot also includes:

[0116] The timing service equipment is connected to the big data acquisition equipment and the parameter capture equipment respectively, and is used to provide various timing service signals required by the big data acquisition equipment and the parameter capture equipment respectively;

[0117] The timing service device is connected to the big data acquisition device and the parameter acquisition device respectively, and is used to provide various timing service signals required by the big data acquisition device and the parameter acquisition device respectively. The timing service device has a built-in quartz oscillation component to provide a reference pulse signal for the construction of timing service signals.

[0118] For example, the timing service device has a built-in quartz oscillator component for providing a reference pulse signal for constructing the timing service signal, including: the reference pulse signal being a square wave of a set frequency.

[0119] Next, various embodiments of the present invention will be further described.

[0120] Optionally, within the above embodiments, in the seamless switching and context synchronization system during the human-machine collaboration process of the customer service robot:

[0121] The intelligent customer service data analysis model is based on set duration thresholds, multiple e-commerce related information of the e-commerce application, duration of time segments, total number of users accessing the e-commerce application in multiple past time segments before the current moment, seamless switching data, and big data trajectory communication content. It intelligently analyzes the number of intelligent robots and human customer service representatives required to achieve seamless switching in the human-machine collaboration process of the customer service robot within the current time segment. The model includes: the number of selected past time segments is positively correlated with the total number of sales product types of the e-commerce application.

[0122] For example, a positive correlation between the number of selected past time segments and the total number of product types sold in the e-commerce application includes: setting the total number of product types sold in the e-commerce application to 20,000 and selecting 10 past time segments; setting the total number of product types sold in the e-commerce application to 30,000 and selecting 15 past time segments; setting the total number of product types sold in the e-commerce application to 40,000 and selecting 20 past time segments; setting the total number of product types sold in the e-commerce application to 50,000 and selecting 25 past time segments, and so on.

[0123] Among them, the positive correlation between the number of selected past time segments and the total number of product types sold in the e-commerce application includes: the selection of information transformation functions to represent the information transformation relationship between the number of selected past time segments and the total number of product types sold in the e-commerce application;

[0124] In the information conversion function, the total number of product types sold by the e-commerce application is set as the input information of the information conversion function, and the number of past time segments positively correlated with the total number of product types sold by the e-commerce application is the output information of the information conversion function.

[0125] And, optionally, within the above embodiments, in the seamless switching and context synchronization system during the human-machine collaboration process of the customer service robot:

[0126] The intelligent customer service data analysis model is based on multiple e-commerce related information of the e-commerce application, the duration of time segments, the total number of users accessing the e-commerce application in multiple past time segments before the current moment, seamless switching data, and big data trajectory communication content. It intelligently analyzes the number of intelligent robots and human customer service representatives required to achieve seamless switching in the human-machine collaboration process of the customer service robot within the current time segment. It also includes: synchronously inputting multiple e-commerce related information of the e-commerce application, the duration of time segments, the total number of users accessing the e-commerce application in multiple past time segments before the current moment, seamless switching data, and big data trajectory communication content into the intelligent customer service data analysis model.

[0127] For example, a PLC device can be selected to synchronously input multiple e-commerce related information of the e-commerce application, the duration of time segments, the total number of users accessing the e-commerce application in multiple past time segments before the current moment, seamless switching data, and big data trajectory communication content into the intelligent customer service data analysis model. The intelligent customer service data analysis model includes multiple input ports.

[0128] The intelligent customer service data analysis model, based on multiple e-commerce related information of the e-commerce application, the duration of time segments, the total number of users accessing the e-commerce application in multiple past time segments before the current moment, seamless switching data, and big data trajectory communication content, intelligently analyzes the number of intelligent robots and human customer service representatives required to achieve seamless switching in the human-machine collaboration process of the customer service robot within the current time segment. This also includes: running the intelligent customer service data analysis model to obtain the number of intelligent robots and human customer service representatives required to achieve seamless switching in the human-machine collaboration process of the customer service robot within the current time segment, as output by the intelligent customer service data analysis model.

[0129] The process of simultaneously inputting multiple e-commerce related information, time segment duration, total number of users accessing the e-commerce application in multiple past time segments before the current moment, seamless switching data, and big data trajectory communication content into the intelligent customer service data analysis model includes: performing binary value conversion on each of the multiple e-commerce related information, time segment duration, total number of users accessing the e-commerce application in multiple past time segments before the current moment, seamless switching data, and big data trajectory communication content before simultaneously inputting them into the intelligent customer service data analysis model;

[0130] Specifically, when performing binary value conversion on various e-commerce related information, time segment duration, total number of users accessing the e-commerce application in multiple past time segments before the current moment, seamless switching data, and big data trajectory communication content of the e-commerce application, the basic information that is itself a binary value does not need to be converted into a binary value again. For example, customer service communication data that is itself in ASCII code does not need to be converted into a binary value since ASCII code is a binary value.

[0131] And among them, running the intelligent customer service data analysis model to obtain the number of intelligent robots and human customer service representatives required to achieve seamless switching in the human-machine collaboration process of the customer service robot within the current time segment, including: the number of intelligent robots and human customer service representatives required to achieve seamless switching in the human-machine collaboration process of the customer service robot within the current time segment is represented in binary numerical form.

[0132] Sixth Embodiment

[0133] Figure 7 This is a flowchart illustrating the steps of a seamless switching and context synchronization method in the human-machine collaboration process of a customer service robot according to the sixth embodiment of the present invention.

[0134] like Figure 7 As shown, the seamless switching and context synchronization method in the human-machine collaboration process of the customer service robot includes the following steps:

[0135] Collect access trajectory information and customer service communication data for each user in each past time segment before the current moment in the e-commerce application, as well as the big data trajectory communication content for that past time segment.

[0136] Here, you can choose to use big data collection devices to collect big data on the access trajectories and customer service interactions of users in each past time segment of the e-commerce application. The collected results are used for the synchronous intelligent analysis of the number of two types of customer service resources to achieve seamless switching in the current time segment.

[0137] Conceptually, the seamless switching between intelligent robots and human customer service is a relative concept, referring to a very short switching time. For example, if the switching time consumed by the intelligent robot and human customer service during a certain time segment is less than or equal to a set time threshold, it is determined that a seamless switching in the human-machine collaboration process of the customer service robot has been achieved within that time segment. Here, the value of the set time threshold is very small.

[0138] Accordingly, the concept of context synchronization in this invention refers to the process of resending the latest communication segment sent by the user to the switched customer service provider (usually a human customer service representative) when switching between the intelligent robot and the human customer service representative.

[0139] Capture seamless switching data for every past time segment before the current moment in the e-commerce application;

[0140] For example, if the current time is 5:00 PM, the current time interval is from 5:00 PM to 5:05 PM, and the multiple past time segments before the current time are multiple past time segments before 5:00 PM, such as 4:55 PM to 5:00 PM, 4:50 PM to 4:55 PM, 4:45 PM to 4:50 PM, 4:40 PM to 4:45 PM, 4:35 PM to 4:40 PM, 4:30 PM to 4:35 PM, 4:25 PM to 4:30 PM, 4:20 PM to 4:25 PM, 4:15 PM to 4:20 PM, and 4:10 PM to 4:15 PM, a total of 10 past time segments;

[0141] Analyze and configure multiple e-commerce association information settings for e-commerce applications;

[0142] Specifically, parsing multiple e-commerce related information of the e-commerce application includes: selecting multiple information collection components to collect multiple e-commerce related information of the e-commerce application respectively. The multiple e-commerce related information of the e-commerce application, like other basic information used for intelligent analysis, can be regarded as a kind of big data to be collected.

[0143] Deep learning is performed on a BP neural network to obtain an intelligent customer service data analysis model;

[0144] For example, numerical simulation mode can be used to test and simulate the model building process of performing deep learning on a BP neural network to obtain an intelligent customer service data analysis model;

[0145] The intelligent customer service data analysis model is based on multiple e-commerce related information of the e-commerce application, the duration of time segments, the total number of users accessing the e-commerce application in multiple past time segments before the current moment, seamless switching data, and big data trajectory communication content. It intelligently analyzes the number of intelligent robots and human customer service representatives required to achieve seamless switching in the human-machine collaboration process of the customer service robot within the current time segment.

[0146] Based on the results of intelligent analysis, configure the e-commerce application with a seamless switching and context synchronization strategy for the current time segmentation.

[0147] Among them, the seamless switching and context synchronization strategy for the e-commerce application segmented based on the intelligent analysis results at the current moment includes: configuring the number of intelligent robots and human customer service representatives obtained by intelligent analysis for the e-commerce application segmented based on the current moment at the current moment, and configuring the context synchronization of the communication segments of intelligent robots and human customer service representatives when switching.

[0148] Among them, capturing seamless switching data of the e-commerce application in each past time segment before the current time includes: capturing the number of intelligent robots and human customer service representatives used to achieve seamless switching in the human-machine collaboration process of the customer service robot in each past time segment before the current time, as the seamless switching data of that past time segment.

[0149] Specifically, for a set e-commerce application, if the switching time consumed when switching between the intelligent robot and the human customer service is less than or equal to the set time threshold within a certain time segment, it is determined that a seamless switching in the human-machine collaboration process of the customer service robot has been achieved within that certain time segment.

[0150] For example, for a set e-commerce application, when the switching time consumed by the intelligent robot and human customer service is less than or equal to a set time threshold in a certain time segment, it is determined that the seamless switching in the human-machine collaboration process of the customer service robot is achieved in a certain time segment, including: the set time threshold is small enough, for example, at the millisecond level;

[0151] The configuration of context synchronization of communication segments when switching between intelligent chatbots and human customer service includes: when switching between intelligent chatbots and human customer service, the latest communication segment sent by the user will be resent to the customer service representative being switched.

[0152] Specifically, when switching between the intelligent robot and the human customer service, the latest communication paragraph sent by the user will be resent to the customer service representative being switched. This can automatically resend the preceding text of the user-customer service communication, improve the speed and efficiency of the switched customer service representative getting into working mode, and also reduce the burden of resending for the user.

[0153] Among them, setting multiple e-commerce related information for the e-commerce application includes setting the number of registered users, the number of registered merchants, the total number of product types sold, the fastest computing speed of the server cluster, and the maximum storage capacity.

[0154] For example, the fastest computing speed using a server cluster can be expressed in terms of operations per second, and the maximum storage capacity using a server cluster can be expressed in terms of MB.

[0155] The current time segment starts at the current moment, and the current time segment and multiple past time segments before the current moment form a complete time interval on the time axis, and the duration of each time segment is equal.

[0156] For example, if the current time is 5:00 PM, the current time interval is from 5:00 PM to 5:05 PM, the multiple past time segments before the current time are the 10 past time segments before 5:00 PM, and the current time segment and the multiple past time segments before the current time form a complete time interval from 4:10 PM to 5:05 PM on the timeline.

[0157] Among them, performing deep learning on the BP neural network to obtain an intelligent customer service data analysis model includes: performing multiple learning operations on the BP neural network more than a preset number of times, so as to obtain the BP neural network after multiple learning operations and use it as the output of the intelligent customer service data analysis model, and the numerical trend of the number of learning operations is consistent with the numerical trend of the number of registered merchants in the set e-commerce application.

[0158] For example, the trend of the number of times the learning is performed is consistent with the trend of the number of registered merchants in the e-commerce application, including: setting the number of registered merchants in the e-commerce application to 5,000 and selecting 1,000 times of learning; setting the number of registered merchants in the e-commerce application to 6,000 and selecting 1,200 times of learning; setting the number of registered merchants in the e-commerce application to 7,000 and selecting 1,400 times of learning; setting the number of registered merchants in the e-commerce application to 8,000 and selecting 1,600 times of learning, and so on.

[0159] In each learning iteration of the BP neural network, the number of intelligent robots and human customer service representatives required for seamless switching in the human-machine collaboration process of the e-commerce application within a certain time segment are taken as the two outputs of the BP neural network. The multiple e-commerce related information of the e-commerce application, the duration of the time segment, the total number of users accessing the e-commerce application in multiple past time segments before the certain time segment, seamless switching data, and big data trajectory communication content are taken as the multiple inputs of the BP neural network to complete this learning iteration of the BP neural network.

[0160] Specifically, the visit duration, visit sequence number, product type number, product return rate, product quantity, and product sales price of each user within each past time segment are used as the visit trajectory information for that user within that past time segment.

[0161] Specifically, the access sequence number of the first product a user visits in a certain past time segment is 1, the access sequence number of the second product a user visits in a certain past time segment is 2, and so on.

[0162] In addition, the binary numerical sequence and the total number of segments formed by sequentially connecting the ASCII values ​​corresponding to each component character of each segment of the interaction between each user and the intelligent robot in each past time segment, and the binary numerical sequence and the total number of segments formed by sequentially connecting the ASCII values ​​corresponding to each component character of each segment of the interaction between each user and the human customer service representative in each past time segment, are used as the customer service interaction data corresponding to the user in that past time segment.

[0163] Furthermore, in the seamless switching and context synchronization system and method for human-machine collaboration in a customer service robot according to the present invention:

[0164] The BP neural network is subjected to multiple learning iterations exceeding a preset number of iterations to obtain a BP neural network after multiple learning iterations, which is then used as the output of the intelligent customer service data analysis model. The numerical trend of the number of learning iterations is consistent with the numerical trend of the number of registered merchants in the set e-commerce application. This includes using a quantity change curve to represent the numerical trend of the number of registered merchants in the set e-commerce application, and using a number change curve to represent the numerical trend of the number of learning iterations.

[0165] The process of performing multiple learning operations on the BP neural network beyond a preset number of times to obtain a BP neural network after multiple learning operations and outputting it as the intelligent customer service data analysis model, and ensuring that the numerical trend of the number of learning operations is consistent with the numerical trend of the number of registered merchants in the e-commerce application, also includes: performing curve length normalization processing on the quantity change curve and the number of learning operations curve respectively to obtain a first change curve and a second change curve with equal curve length, and the first change curve and the second change curve completely overlap.

[0166] For example, performing multiple learning iterations on the BP neural network beyond a preset number of iterations to obtain a BP neural network after multiple learning iterations is used as the output of the intelligent customer service data analysis model. Furthermore, the numerical trend of the number of learning iterations is consistent with the numerical trend of the number of registered merchants in the set e-commerce application. This also includes: the BP neural network selecting the Sigmoid function as the activation function.

[0167] The above description is merely a specific implementation of the embodiments of this application, but the protection scope of the embodiments of this application is not limited thereto. Any changes or substitutions within the technical scope disclosed in the embodiments of this application should be covered within the protection scope of the embodiments of this application. Therefore, the protection scope of the embodiments of this application should be determined by the protection scope of the claims.

Claims

1. A seamless switching and context synchronization system for human-machine collaboration in a customer service robot, characterized in that, The system includes: Big data collection equipment is used to collect access trajectory information and customer service communication data of each user in each past time segment before the current moment of a set e-commerce application, so as to serve as the big data trajectory communication content of that past time segment. A parameter capture device used to capture seamless switching data for each past time segment before the current moment in a specified e-commerce application; Information parsing equipment, used to parse multiple e-commerce related information for setting up e-commerce applications; A continuous reconfiguration device is used to perform deep learning on a BP neural network to obtain an intelligent customer service data analysis model. The data analysis equipment is connected to the big data acquisition equipment, parameter capture equipment, information parsing equipment, and continuous reconstruction equipment, respectively. It is used to use the intelligent customer service data analysis model to analyze multiple e-commerce related information of the set e-commerce application, the duration of time segments, the total number of users accessing the set e-commerce application in multiple past time segments before the current moment, seamless switching data, and big data trajectory communication content to achieve the number of intelligent robots and human customer service representatives required for seamless switching in the human-machine collaboration process of the customer service robot within the current moment segment. Dynamically configure devices, connect to data analysis devices, and configure seamless switching and context synchronization strategies for e-commerce applications based on intelligent analysis results at the current moment, segmented into current time periods. The seamless switching and context synchronization strategy for the e-commerce application segmented based on the results of intelligent analysis includes: configuring the number of intelligent robots and human customer service representatives obtained from intelligent analysis for the e-commerce application segmented based on the results of intelligent analysis, and configuring the context synchronization of the communication segments when switching between intelligent robots and human customer service representatives. Among them, capturing seamless switching data of the e-commerce application in each past time segment before the current time includes: capturing the number of intelligent robots and human customer service representatives used to achieve seamless switching in the human-machine collaboration process of the customer service robot in each past time segment before the current time, as the seamless switching data of that past time segment. Specifically, for a set e-commerce application, if the switching time consumed when switching between the intelligent robot and the human customer service is less than or equal to the set time threshold within a certain time segment, it is determined that a seamless switching in the human-machine collaboration process of the customer service robot has been achieved within that certain time segment. The configuration of context synchronization of communication segments when switching between intelligent chatbots and human customer service includes: when switching between intelligent chatbots and human customer service, the latest communication segment sent by the user will be resent to the customer service representative being switched. Among them, setting multiple e-commerce related information for the e-commerce application includes setting the number of registered users, the number of registered merchants, the total number of product types sold, the fastest computing speed of the server cluster, and the maximum storage capacity. The current time segment starts at the current moment, and the current time segment and multiple past time segments before the current moment form a complete time interval on the time axis, and the duration of each time segment is equal.

2. The seamless switching and context synchronization system for human-machine collaboration in customer service robots as described in claim 1, characterized in that: Performing deep learning on a BP neural network to obtain an intelligent customer service data analysis model includes: performing multiple learning operations on the BP neural network beyond a preset number of times, so as to obtain a BP neural network after multiple learning operations and use it as the output of the intelligent customer service data analysis model, and the numerical trend of the number of learning operations is consistent with the numerical trend of the number of registered merchants in the set e-commerce application. In each learning iteration of the BP neural network, the number of intelligent robots and human customer service representatives required for seamless switching in the human-machine collaboration process of the e-commerce application within a certain time segment are taken as the two outputs of the BP neural network. The multiple e-commerce related information of the e-commerce application, the duration of the time segment, the total number of users accessing the e-commerce application in multiple past time segments before the certain time segment, seamless switching data, and big data trajectory communication content are taken as the multiple inputs of the BP neural network to complete this learning iteration of the BP neural network. Specifically, the visit duration, visit sequence number, product type number, product return rate, product quantity, and product sales price of each user within each past time segment are used as the visit trajectory information for that user within that past time segment. Specifically, the binary numerical sequence and the total number of segments formed by sequentially connecting the ASCII values ​​corresponding to each character in each segment of the interaction between each user and the intelligent robot within each past time segment, and the binary numerical sequence and the total number of segments formed by sequentially connecting the ASCII values ​​corresponding to each character in each segment of the interaction between each user and the human customer service representative within each past time segment, are used as the customer service interaction data for that user within that past time segment.

3. The seamless switching and context synchronization system for human-machine collaboration in customer service robots as described in claim 2, characterized in that, The system also includes: The real-time display device is set in the user client device running the set e-commerce application and connected to the data analysis device. It is used to receive the number of intelligent robots and human customer service representatives required for seamless switching in the human-machine collaboration process of the customer service robot within the current time segment. The instant display device is also used to simultaneously display the number of intelligent robots and human customer service representatives received.

4. The seamless switching and context synchronization system for human-machine collaboration in customer service robots as described in claim 2, characterized in that, The system also includes: A model storage device, connected to a continuous reconstruction device, is used to receive intelligent customer service data analysis models and store the intelligent customer service data analysis models. The process of receiving and storing intelligent customer service data analysis models includes: storing the intelligent customer service data analysis models by storing various model parameters of the intelligent customer service data analysis models.

5. The seamless switching and context synchronization system for human-machine collaboration in a customer service robot as described in claim 2, characterized in that, The system also includes: The timing service equipment is connected to the big data acquisition equipment and the parameter capture equipment respectively, and is used to provide various timing service signals required by the big data acquisition equipment and the parameter capture equipment respectively; The timing service device is connected to the big data acquisition device and the parameter acquisition device respectively, and is used to provide various timing service signals required by the big data acquisition device and the parameter acquisition device respectively. The timing service device has a built-in quartz oscillator component to provide a reference pulse signal for the construction of timing service signals.

6. The seamless switching and context synchronization system for human-machine collaboration in customer service robots as described in claim 2, characterized in that: The intelligent customer service data analysis model is based on set duration thresholds, multiple e-commerce related information of the e-commerce application, duration of time segments, total number of users accessing the e-commerce application in multiple past time segments before the current moment, seamless switching data, and big data trajectory communication content. It intelligently analyzes the number of intelligent robots and human customer service representatives required to achieve seamless switching in the human-machine collaboration process of the customer service robot within the current time segment. The model includes: the number of selected past time segments is positively correlated with the total number of sales product types of the e-commerce application. Among them, the positive correlation between the number of selected past time segments and the total number of product types sold in the e-commerce application includes: the selection of information transformation functions to represent the information transformation relationship between the number of selected past time segments and the total number of product types sold in the e-commerce application; In the information conversion function, the total number of product types sold by the e-commerce application is set as the input information of the information conversion function, and the number of past time segments that are positively correlated with the total number of product types sold by the e-commerce application is the output information of the information conversion function.

7. A method for seamless switching and context synchronization during human-machine collaboration in a customer service robot, characterized in that, The method includes: The method is based on the seamless switching and context synchronization system in the human-machine collaboration process of the customer service robot as described in any one of claims 1-6; Collect access trajectory information and customer service communication data for each user in each past time segment before the current moment in the e-commerce application, as well as the big data trajectory communication content for that past time segment. Capture seamless switching data for every past time segment before the current moment in the e-commerce application; Analyze and configure multiple e-commerce association information settings for e-commerce applications; Deep learning is performed on a BP neural network to obtain an intelligent customer service data analysis model; The intelligent customer service data analysis model is based on multiple e-commerce related information of the e-commerce application, the duration of time segments, the total number of users accessing the e-commerce application in multiple past time segments before the current moment, seamless switching data, and big data trajectory communication content. It intelligently analyzes the number of intelligent robots and human customer service representatives required to achieve seamless switching in the human-machine collaboration process of the customer service robot within the current time segment. Based on intelligent analysis results, the e-commerce application is configured with a seamless switching and context synchronization strategy for the current time segment.

Images