Customer contact method and device suitable for robot customer service

Abstract

The invention discloses a customer contact method and device suitable for robot customer service, relates to the technical field of artificial intelligence and can also be used in the financial field, and the method comprises the steps: screening out a target customer according to a service identifier of a customer and a pre-generated service feature extraction model; wherein the target customer is a customer who is about to provide service for the robot customer service; determining an optimal path between the target customer and the robot customer service according to the position of the target customer and the position of the robot customer service and in combination with at least one of the positions of customers around the target customer and the positions of customers around the robot customer service; and performing customer contact service on the target customer according to the optimal path. According to the invention, the robot customer service can be moved to the customer position needing to be served, and one-stop service is provided, so that the service efficiency is greatly improved, and the customer service experience is improved.

Description

Technical Field

[0001] The present application belongs to the field of artificial intelligence technology, and specifically relates to a customer contact method and device suitable for robot customer service. Background Art

[0002] Existing business processing methods do not pay enough attention to special groups. Although most companies have introduced self-service equipment, such as ATMs and self-service terminals, these devices are still difficult to operate for special groups such as the elderly and the visually impaired. New business processing methods, such as online banking and mobile banking, are convenient and fast, but for this group of people, they may not be able to accept or use them proficiently. Summary of the invention

[0003] The present invention can be used in the technical field of application of artificial intelligence in finance, and can also be used in any field other than the financial field.

[0004] One object of the present invention is to provide a customer contact method suitable for robot customer service, which can enable the robot customer service to move to the location of the customer who needs service and provide one-stop service, thereby greatly improving service efficiency and enhancing customer service experience.

[0005] Another object of the present invention is to provide a customer contact device suitable for robot customer service. Another object of the present invention is to provide an electronic device, the electronic device comprising a memory and a processor, the memory storing a computer program, and the processor implementing the steps of the above-mentioned customer contact method suitable for robot customer service when executing the computer program. Another object of the present invention is to provide a readable medium, on which a computer program is stored, and when the computer program is executed by the processor, the steps of the above-mentioned customer contact method suitable for robot customer service are implemented.

[0006] In order to solve the technical problems in the background technology of this application, the present invention provides the following technical solutions:

[0007] In a first aspect, the present invention provides a customer contact method applicable to robot customer service, comprising:

[0008] Filter out target customers according to the customer's service identifier and the pre-generated service feature extraction model; wherein the target customers are customers to be served by the robot customer service;

[0009] Determine an optimal path between the target customer and the robot customer service according to the location of the target customer, the location of the robot customer service, and at least one of the locations of customers around the target customer and the locations of customers around the robot customer service;

[0010] Provide customer service to the target customer according to the optimal path.

[0011] In some embodiments of the present invention, the step of generating the service feature extraction model includes:

[0012] Acquire image data corresponding to the historical data of the service identifier;

[0013] The initial model of the service feature extraction model is trained according to the pixel values ​​of the image data to generate the service feature extraction model; wherein the initial model is a convolutional recurrent neural network.

[0014] In some embodiments of the present invention, training the initial model of the service feature extraction model according to the pixel values ​​of the image data includes:

[0015] Inputting the pixel value into a convolutional layer in the initial model to extract features of the service identifier;

[0016] Inputting the features of the service identifier into the recurrent layer in the initial model to extract sequence data of the service identifier; wherein the sequence data is used to characterize the relationship between the features of the service identifier and its context;

[0017] Inputting the features and the sequence data into the connection layer of the initial model to generate a character probability distribution corresponding to the image data;

[0018] The initial model is forward propagated and / or backward propagated according to the character probability distribution and a preset threshold.

[0019] In some embodiments of the present invention, the step of screening out target customers according to the customer's service identifier and a pre-generated service feature extraction model includes:

[0020] Converting the image data corresponding to the service identifier into a grayscale image;

[0021] Performing threshold processing on the grayscale image to generate a threshold processing result;

[0022] Inputting the threshold processing result into the service feature extraction model to generate the character probability distribution;

[0023] Determining a text sequence corresponding to the service identifier according to the character probability distribution;

[0024] The target customers are screened out based on the text sequence and the local customer contact service scope.

[0025] In some embodiments of the present invention, the service identifier is a queue number of a service that the customer wants to obtain; the customer contact method further includes:

[0026] After the customer obtains the queue number and the service the customer wants to obtain falls within the scope of the local touch customer service, the customer's location is monitored in real time.

[0027] In some embodiments of the present invention, when there are no customers around the robot customer service and there are no customers around the target customer, determining the optimal path between the target customer and the robot customer service according to the location of the target customer and the location of the robot customer service includes:

[0028] Generate a grid map corresponding to the scene where the robot customer service and the target customer are located;

[0029] Determine the movement cost of each grid in the grid map according to the location of the target customer and the location of the robot customer service; wherein the movement cost includes an actual cost and a predicted cost; the actual cost is the path loss of the robot customer service moving to the current grid, and the predicted cost is the path loss of the robot moving from the current grid to the location of the target customer;

[0030] An optimal path is determined according to the movement cost of each grid.

[0031] In some embodiments of the present invention, when there are customers around the robot customer service and / or there are customers around the target customer, determining the optimal path between the target customer and the robot customer service according to the location of the target customer, the location of the robot customer service, and combining at least one of the locations of customers around the target customer and the locations of customers around the robot customer service includes:

[0032] Generate an initial path according to the movement cost of each grid;

[0033] Generate at least one alternative path based on the initial path and in combination with at least one of the locations of customers around the target customer and the locations of customers around the robot customer service;

[0034] The path is adjusted in real time in the initial path and the at least one alternative path according to the current locations of the customers around the target customer and / or the current locations of the customers around the robot customer service to determine the optimal path.

[0035] In a second aspect, the present invention provides a customer contact device suitable for robot customer service, the device comprising:

[0036] A target customer screening module is used to screen out target customers according to the customer's service identification and a pre-generated service feature extraction model; wherein the target customer is a customer to be served by the robot customer service;

[0037] an optimal path determination module, configured to determine an optimal path between the target customer and the robot customer service according to the location of the target customer, the location of the robot customer service, and at least one of the locations of customers around the target customer and the locations of customers around the robot customer service;

[0038] The customer contact service module is used to provide customer contact service to the target customer according to the optimal path.

[0039] In some embodiments of the present invention, a customer contact device suitable for robot customer service further includes:

[0040] A model generation module is used to generate the service feature extraction model, and the model generation module includes:

[0041] An image data acquisition unit, used to acquire image data corresponding to the historical data of the service identifier;

[0042] A model generation unit is used to train the initial model of the service feature extraction model according to the pixel values ​​of the image data to generate the service feature extraction model; wherein the initial model is a convolutional recurrent neural network.

[0043] In some embodiments of the present invention, the model generation unit includes:

[0044] A feature extraction unit, used for inputting the pixel value into the convolution layer in the initial model to extract the features of the service identifier;

[0045] A sequence data extraction unit, used for inputting the features of the service identifier into the recurrent layer in the initial model to extract the sequence data of the service identifier; wherein the sequence data is used to characterize the relationship between the features of the service identifier and its context;

[0046] A character distribution generating unit, used for inputting the features and the sequence data into the connection layer of the initial model to generate a character probability distribution corresponding to the image data;

[0047] The model training unit is used to perform forward propagation and / or backward propagation on the initial model according to the character probability distribution and a preset threshold.

[0048] In some embodiments of the present invention, the target customer screening module includes:

[0049] A grayscale image generating unit, used for converting the image data corresponding to the service identifier into a grayscale image;

[0050] A threshold processing result generating unit, used for performing threshold processing on the grayscale image to generate a threshold processing result;

[0051] A probability distribution generating unit, used for inputting the threshold processing result into the service feature extraction model to generate the character probability distribution;

[0052] A text sequence determination unit, used to determine the text sequence corresponding to the service identifier according to the character probability distribution;

[0053] The target customer screening unit is used to screen out the target customers according to the text sequence and the local customer contact service scope.

[0054] In some embodiments of the present invention, the service identifier is a queue number of a service that the customer wants to obtain; the customer contact device further includes:

[0055] The real-time location monitoring module is used to monitor the customer's location in real time after the customer obtains the queue number and the service the customer wants to obtain falls within the scope of the local contact customer service.

[0056] In some embodiments of the present invention, when there are no customers around the robot customer service and there are no customers around the target customer, the optimal path determination module includes:

[0057] A grid map generating unit, used to generate a grid map corresponding to the scene where the robot customer service and the target customer are located;

[0058] A mobile cost determination unit, configured to determine the mobile cost of each grid in the grid map according to the location of the target customer and the location of the robot customer service; wherein the mobile cost includes an actual cost and a predicted cost; the actual cost is the path loss of the robot customer service moving to the current grid, and the predicted cost is the path loss of the robot moving from the current grid to the location of the target customer;

[0059] The optimal path determination unit is used to determine the optimal path according to the movement cost of each grid.

[0060] In some embodiments of the present invention, when there are customers around the robot customer service and / or there are customers around the target customer, the optimal path determination module includes:

[0061] An initial path generating unit, used for generating an initial path according to the movement cost of each grid;

[0062] An alternative path generating unit, configured to generate at least one alternative path according to the locations of customers around the target customer, or the locations of customers around the robot customer service and the initial path;

[0063] A path adjustment unit is used to adjust the path in the initial path and the at least one alternative path in real time according to the current locations of customers around the target customer and / or the current locations of customers around the robot customer service to determine the optimal path.

[0064] In a third aspect, the present invention provides a computer program product, comprising a computer program / instructions, which, when executed by a processor, implement the steps of a customer contact method suitable for robot customer service.

[0065] In a fourth aspect, the present invention provides an electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the processor implements the steps of a customer contact method applicable to robot customer service.

[0066] In a fifth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of a customer contact method applicable to robot customer service.

[0067] From the above description, it can be seen that an embodiment of the present invention provides a customer contact method and device suitable for robot customer service, and the corresponding method includes: first, screening out target customers based on the customer's service identification and a pre-generated service feature extraction model; wherein the target customer is a customer to be served by the robot customer service; then, determining the optimal path between the target customer and the robot customer service based on the target customer's location, the robot customer service location, or the locations of customers around the target customer, or the locations of customers around the robot customer service; finally, providing customer contact service to the target customer according to the optimal path.

[0068] The present invention provides a customer contact method and device suitable for robot customer service, so that the robot customer service can automatically move to the location of the customer who needs service and provide one-stop service, thereby greatly improving service efficiency and enhancing service experience. At the same time, the mobility of the robot customer service also solves the accessibility problem of fixed self-service equipment. In addition, the robot customer service can provide a series of services from simple to complex, making the service experience more coherent and reducing the waiting time of customers. This mobile service robot customer service provides a new service mode for the service industry, which can further improve service efficiency and customer satisfaction. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0070] Figure 1 Schematic diagram of a customer contact method applicable to robot customer service in an embodiment of the present invention Figure 1 ;

[0071] Figure 2 A schematic diagram of a customer contact method applicable to robot customer service in an embodiment of the present invention Figure 2 ;

[0072] Figure 3 It is a flowchart of step 400 of a customer contact method applicable to robot customer service in an embodiment of the present invention;

[0073] Figure 4 4 is a flow chart of step 402 of a customer contact method applicable to robot customer service in an embodiment of the present invention;

[0074] Figure 5 It is a flowchart of step 100 of a customer contact method applicable to robot customer service in an embodiment of the present invention;

[0075] Figure 6 A schematic diagram of a customer contact method applicable to robot customer service in an embodiment of the present invention Figure 3 ;

[0076] Figure 7 Schematic diagram of step 200 of a customer contact method applicable to robot customer service in an embodiment of the present invention Figure 1 ;

[0077] Figure 8 Schematic diagram of step 200 of a customer contact method applicable to robot customer service in an embodiment of the present invention Figure 2 ;

[0078] Fig. 9 It is a block diagram of a customer contact system suitable for robot customer service in a specific embodiment of the present invention;

[0079] Fig.10 It is a flowchart of a customer contact method applicable to robot customer service in a specific embodiment of the present invention;

[0080] Fig.11 A block of a customer contact device suitable for robot customer service in an embodiment of the present invention Figure 1 ;

[0081] Fig.12 A block of a customer contact device suitable for robot customer service in an embodiment of the present invention Figure 2 ;

[0082] Fig.13is a block diagram of a model generation module 40 in an embodiment of the present invention;

[0083] Fig.14 is a block diagram of a model generation unit 40b in an embodiment of the present invention;

[0084] Fig.15 is a block diagram of a target customer screening module 10 in an embodiment of the present invention;

[0085] Fig.16 A block of a customer contact device suitable for robot customer service in an embodiment of the present invention Figure 3 ;

[0086] Fig.17 is a block diagram of the optimal path determination module 20 in an embodiment of the present invention;

[0087] Fig.18 is a block diagram of the optimal path determination module 20 in an embodiment of the present invention;

[0088] Fig.19 It is a schematic diagram of the structure of an electronic device in an embodiment of the present invention. DETAILED DESCRIPTION

[0089] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0090] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0091] It should be noted that the terms "including" and "having" in the specification and claims of the present application and the above-mentioned drawings and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices. In the absence of conflict, the embodiments in the present application and the features in the embodiments may be combined with each other. The present application will be described in detail below with reference to the drawings and in combination with the embodiments.

[0092] The acquisition, storage, use, and processing of data in this application's technical solution are in compliance with the relevant provisions of laws and regulations. Specifically:

[0093] The information collected in the technical solution of this application is information and data authorized by the user or fully authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of relevant data comply with the relevant laws, regulations and standards of relevant countries and regions, take necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0094] Provide users with corresponding operation entrances for them to choose to agree or reject the automated decision-making results; if the user chooses to reject, the expert decision-making process will be entered.

[0095] The embodiment of the present invention provides a specific implementation of a customer contact method applicable to a robot customer service, see Figure 1 , the method specifically includes the following contents:

[0096] Step 100: Screening out target customers according to the customer's service identifier and the pre-generated service feature extraction model; wherein the target customer is the customer to be served by the robot customer service;

[0097] Step 200: Determine an optimal path between the target customer and the robot customer service according to the location of the target customer, the location of the robot customer service, and at least one of the locations of customers around the target customer and the locations of customers around the robot customer service;

[0098] Step 300: Provide customer service to the target customer according to the optimal path.

[0099] From the above description, it can be seen that an embodiment of the present invention provides a customer contact method suitable for robot customer service, including: first, screening out target customers based on the customer's service identification and a pre-generated service feature extraction model; wherein the target customer is a customer to be served by the robot customer service; then, determining the optimal path between the target customer and the robot customer service based on the target customer's location, the robot customer service location, or the locations of customers around the target customer, or the locations of customers around the robot customer service; finally, providing customer contact service to the target customer according to the optimal path.

[0100] The present invention enables the robot customer service to automatically move to the customer's location and provide fast and efficient one-stop service. By optimizing the service process, the business processing efficiency and service quality are greatly improved. At the same time, the mobility of the robot customer service makes up for the accessibility problem of fixed self-service equipment, so that all customers, including those with physical disabilities or unfamiliar with the location of the equipment, can enjoy convenient services. In addition, this robot customer service can provide a series of coherent services, reduce the customer's waiting time, and make the service experience smoother.

[0101] It can be understood that the service identifier in step 100 is used to indicate that the service that the corresponding customer wants to enjoy is a service that can be provided by the robot customer service. Preferably, the service identifier can be the queue number of the customer under the corresponding service.

[0102] For step 200, there are at least two cases:

[0103] The first one is that when there are no customers (other customers) around the robot customer service and there are no customers (other customers) around the target customer, it is only necessary to determine the optimal path between the target customer and the robot customer. It should be understood that the route may change due to changes in the target customer's location.

[0104] The second type is when there are customers around the robot customer service and there are customers around the target customer, it is necessary to consider the impact of position changes of the customers around the robot and the position changes of the customers around the target customer on the optimal path. For example, after the position changes, the customers around the robot and the customers around the target customer are located on the previously determined optimal path, so at least one alternative path needs to be determined.

[0105] In addition, there may be customers only around the robot customer service, but no customers around the target customers; or there may be customers only around the target customers, but no customers around the robot customer service. In this case, only one of the situations needs to be considered to affect the optimal path.

[0106] Step 300 is to provide the target customer with the service he wants to enjoy based on the determination of the optimal path, and the service is also within the service scope that can be provided by the local robot customer service. This method not only saves the customer's time, but also saves the labor cost of the local end that provides the service.

[0107] In some embodiments of the present invention, see Figure 2 , a customer contact method suitable for robot customer service, further comprising:

[0108] Step 400: Generate the service feature extraction model;

[0109] It can be understood that the service feature extraction model is used to extract features in the service identifier to determine whether the service corresponding to the feature (the service that the customer wants to enjoy) is within the scope of services that can be provided by the local robot customer.

[0110] Next, see Figure 3 , step 400 comprises:

[0111] Step 401: Acquire image data corresponding to the historical data of the service identifier;

[0112] Specifically, when step 401 is implemented, the local robot customer service can obtain the image data corresponding to the historical data of the service identifier through its own image acquisition device.

[0113] Step 402: Train the initial model of the service feature extraction model according to the pixel values ​​of the image data to generate the service feature extraction model; wherein the initial model is a convolutional recurrent neural network.

[0114] First, provide each image with the correct text labels. The labels should exactly match the text in the image, including order and capitalization. Convert the image to grayscale or normalize it to reduce computational complexity, and apply image enhancement techniques such as binarization, denoising, contrast enhancement, etc. to improve the visibility of the text.

[0115] Use multiple convolutional layers and pooling layers of the initial model to extract features of the image. These layers can gradually extract text features and reduce the spatial size of the feature map. After the convolutional layer, use one or more recurrent layers (such as LSTM or GRU) to process sequence information to capture dependencies within the sequence, such as the order and structure of the text. Use a fully connected layer to map the output of the recurrent layer to a character set. Use a softmax activation function on the output to calculate the probability distribution of each character. Preferably, a CTC loss function can be used to train the network.

[0116] In some embodiments of the present invention, see Figure 4 , step 402 includes:

[0117] Step 4021: inputting the pixel value into the convolution layer in the initial model to extract the features of the service logo;

[0118] Taking into account the implementation scenario of step 4021 (service logos often contain fewer words), the convolutional layer in step 4021 contains 2 convolutional layers and 3 pooling layers to gradually reduce the spatial dimension and increase the depth of the feature map.

[0119] Step 4022: Input the features of the service identifier into the recurrent layer in the initial model to extract sequence data of the service identifier; wherein the sequence data is used to characterize the relationship between the features of the service identifier and its context;

[0120] Specifically, a recurrent layer, such as LSTM or GRU, is added after the convolutional layer to process sequence data. Preferably, the components of LSTM include: Input gate: controls whether to update the cell state. Forget gate: controls whether to delete the previous state information. Cell state: stores long-term information. Output gate: controls the flow of information from the cell state to the hidden state. LSTM can flexibly choose to remember and forget information through the gating mechanism, so it is more effective in processing long sequences.

[0121] GRU combines the input gate and forget gate in LSTM, and uses the update gate to control the update of information. GRU has only two gate units: Update gate: controls the weight between the old state and the new state. Reset gate: controls the forgetting of the previous hidden state.

[0122] Step 4023: inputting the features and the sequence data into the connection layer of the initial model to generate a character probability distribution corresponding to the image data;

[0123] Specifically, a fully connected layer is used to map the output of the recurrent layer to the character set, and a softmax activation function is used on the output to calculate the probability distribution of each character (the probability that the next possible character is each possible character).

[0124] Step 4024: Perform forward propagation and / or backward propagation on the initial model according to the character probability distribution and a preset threshold.

[0125] The forward propagation passes the input data through the network until it reaches the output layer. During this process, the output of each layer will be used as the input of the next layer. The input data is passed to one or more hidden layers. In each layer, the input data is multiplied by the weight matrix of that layer, a bias is added, and then processed through a nonlinear activation function. This process can be expressed mathematically as:

[0126] a [l =g [l](W [l] a [l-1] +b [l] )]

[0127] Among them, a [l] is the activation value of layer l, W [l] and b [l] are the weight and bias of the lth layer, g [l] is the activation function. The output of the last hidden layer is passed to the output layer, which is usually designed to produce the final output of the network, such as class probabilities in classification tasks or continuous values ​​in regression tasks.

[0128] Back propagation is used to calculate the gradient of the loss function with respect to the network parameters and update the parameters accordingly to minimize the loss function. The purpose of this process is to optimize the network weights for better prediction. Specifically, the steps of back propagation are as follows:

[0129] First, we calculate the loss function, which is a measure of the difference between the network's predicted output and the actual label. Common loss functions include mean squared error, cross entropy, etc. The gradient of the loss function with respect to each parameter is calculated layer by layer using the chain rule. This involves going from the output layer to the input layer, calculating the gradient of the weights and biases of each layer. The gradient calculation formula is:

[0130]

[0131] Where L is the loss function. Next, use gradient descent or other optimization algorithms to update the weights and biases of the network. The update formula is:

[0132]

[0133] Among them, α is the learning rate, which is used to control the step size of weight adjustment.

[0134] In some embodiments of the present invention, see Figure 5 , step 100 comprises:

[0135] Step 101: converting the image data corresponding to the service logo into a grayscale image;

[0136] Step 102: performing threshold processing on the grayscale image to generate a threshold processing result;

[0137] Specifically, by traversing all possible thresholds, the intra-class variance and inter-class variance generated when each threshold divides the image into foreground (target) and background are calculated. The threshold that maximizes the inter-class variance is selected to maximize the difference between the foreground and background. Specifically:

[0138] First, we need to calculate the grayscale histogram of the entire image, that is, the frequency of each grayscale. Next, calculate the average grayscale of the entire image. Traverse all possible thresholds: For each possible threshold (t), divide the histogram into two parts (foreground and background). Calculate the pixel weights w0(t) and w1(t) of the foreground and background respectively. Calculate the average grayscale μ0(t) and μ1(t) of the foreground and background. Calculate the inter-class variance:

[0139]

[0140] here is the between-class variance, which is used to find Maximize t. Finally, choose the best threshold: threshold t makes The largest value is selected as the optimal threshold.

[0141] Step 103: inputting the threshold processing result into the service feature extraction model to generate the character probability distribution;

[0142] Preferably, CTC decoding is used to generate the most likely character sequence from a continuous probability distribution. Preferably, spaces between characters are processed and repeated characters are merged.

[0143] Step 104: determining a text sequence corresponding to the service identifier according to the character probability distribution;

[0144] Step 105: Filter out the target customers according to the text sequence and the local customer contact service scope.

[0145] In some embodiments of the present invention, the service identifier is the queue number of the service that the customer wants to obtain; see Figure 6 , a customer contact method suitable for robot customer service, further comprising:

[0146] Step 500: After the customer obtains the queue number and the service the customer wants to obtain falls within the scope of the local contact service, the customer's location is monitored in real time.

[0147] It is understandable that when there are multiple target customers and the number of currently available robot customer service staff is 0, the target customers need to be placed in the waiting sequence, but the customer's location needs to be monitored in real time to prepare for determining the optimal path later.

[0148] In some embodiments of the present invention, see Figure 7 , when there are no customers around the robot customer service and there are no customers around the target customer, determining the optimal path between the target customer and the robot customer service according to the location of the target customer and the location of the robot customer service, including:

[0149] Step 201: Generate a grid map corresponding to the scene where the robot customer service and the target customer are located;

[0150] Step 202: Determine the movement cost of each grid in the grid map according to the location of the target customer and the location of the robot customer service; wherein the movement cost includes an actual cost and a predicted cost; the actual cost is the path loss of the robot customer service moving to the current grid, and the predicted cost is the path loss of the robot moving from the current grid to the location of the target customer;

[0151] Step 203: Determine the optimal path according to the movement cost of each grid.

[0152] In step 201 to step 203, two lists need to be set: one is the Open List, which stores nodes that have been discovered but not yet processed (i.e., nodes whose neighbors have been checked), and the other is the Closed List, which stores nodes that have been processed. Specifically, add the starting node to the Open List. Repeat the following steps until the Open List is empty:

[0153] a. Find the node with the lowest move cost value from the open list and call it the current node.

[0154] b. Move the current node from the open list to the closed list.

[0155] c. If the current node is the target node, then the path has been found and the final path can be constructed by backtracking the path.

[0156] d. Otherwise, for each neighbor of the current node:

[0157] i. If it is not in the open list and closed list, calculate its move cost value and add it to the open list.

[0158] ii. If it is already in the open list or closed list, check whether it is better to get there through the current node, that is, check whether the movement cost is lower. If so, update the movement cost value of that node and reassign its parent node as the current node.

[0159] If the open list becomes empty, there is no path to reach the target node and the algorithm fails.

[0160] In some embodiments of the present invention, see Figure 8, when there are customers around the robot customer service and / or there are customers around the target customer, determining the optimal path between the target customer and the robot customer service according to the location of the target customer, the location of the robot customer service, and combining at least one of the locations of customers around the target customer and the locations of customers around the robot customer service, including:

[0161] Step 204: generating an initial path according to the movement cost of each grid;

[0162] Step 205: Generate at least one alternative path based on the initial path and in combination with at least one of the locations of customers around the target customer and the locations of customers around the robot customer service;

[0163] The method for generating the candidate path may refer to step 201 to step 203 .

[0164] Step 206: According to the current locations of customers around the target customer and / or the current locations of customers around the robot customer service, adjust the path in the initial path and the at least one alternative path in real time to determine the optimal path.

[0165] In order to further illustrate the solution, the present invention also takes a bank branch as an example to provide a specific implementation method of a customer contact method suitable for robot customer service, which specifically includes the following contents.

[0166] Bank branch business has undergone an unprecedented transformation in the past few decades. From the initial cumbersome paper bookkeeping to the current electronic and automated services, the way banking business is handled has undergone earth-shaking changes. With the development of information technology, the core business processing system (CBS) has become the infrastructure for banking business processing, while online banking and mobile banking have truly brought banking business into the digital age.

[0167] However, there are some problems with the existing banking business processing methods. For some complex businesses, such as account opening, account cancellation, reporting lost cards, etc., manual participation is still required, and the efficiency of manual service is limited. Customers usually need to wait for a long time in the bank lobby. The service quality is affected by individual differences in teller business ability, service attitude, work status, etc., resulting in uneven service quality. The shortcomings of the existing banking business processing methods are mainly reflected in the following aspects:

[0168] (1) Service efficiency. Service efficiency is a major problem. Even with the assistance of self-service devices, customers may still spend too much time looking for services or equipment. When the number of customers increases, especially during peak hours, the service efficiency of the manual counter may drop significantly. At this time, the self-service device may not be able to effectively relieve the pressure on the manual counter.

[0169] (2) Unstable service quality. Since service quality is affected by the service personnel’s skills, attitude, and working status, different service results and user experience may occur, which leads to unstable service quality.

[0170] (3) Service accessibility is also an important issue. In a bank lobby, self-service devices are usually located in fixed locations, while customers may come from anywhere in the lobby. For customers with physical limitations or unfamiliarity with the location of the device, this may cause inconvenience and affect their service experience.

[0171] (4) The connection between manual service and self-service cannot be ignored. For some complex businesses, customers may need to switch between manual service and self-service, which may lead to an inconsistent service experience and increase customer waiting time.

[0172] The present invention aims to solve several core problems in the existing banking business processing methods: low efficiency of manual service, unstable service quality, poor service accessibility, and poor connection between manual service and self-service. Based on this, the present invention provides a customer contact method suitable for robot customer service based on area recognition, target positioning, path planning and automatic navigation.

[0173] First, the present invention provides a customer contact system suitable for robot customer service, see Fig. 9 The system includes: a computer vision module, a target recognition module, a path planning module, an automatic navigation module and a service interaction module. Specifically:

[0174] Computer vision module: The camera or other image capture device installed on the robot customer service obtains image data of the surrounding environment in real time. The image data is sent to the computer vision processing unit, which processes the image using computer vision algorithms, identifies the ground and obstacles, and models the working environment of the robot customer service. Obstacle information is obtained through the collected images to create a two-dimensional grid map. Each cell may represent a walkable open space or an inaccessible obstacle. Then, the latest images are continuously processed to update the two-dimensional grid map in order to identify dynamically changing obstacles.

[0175] Target recognition module: used for data collection and model training, customer vital signs extraction and feature recognition.

[0176] Data collection and model training: First, the training data needs to be collected. This includes a large number of images of people holding call number papers, as well as images of non-target customers, such as people who do not hold call number papers or whose hands are not in the image. After collecting these images, you can start training the convolutional neural network (CNN) model. The pixel values ​​of the image are used as input, and the parameters of the CNN model are continuously adjusted to reduce the prediction error of the model through forward propagation and back propagation. In each training iteration, the parameters of the model are updated to improve the prediction accuracy of the model on the training data. After a large number of iterations, the model training is completed.

[0177] Customer feature acquisition: The robot customer service needs to use its built-in camera or other image acquisition equipment. The robot customer service uses these devices to scan the environment in all directions and capture the image information of all people present. In this image information, the robot customer service uses algorithms to pay special attention to and extract the customer's hands and the area of ​​the call number paper.

[0178] Feature recognition: The robot customer service will input the acquired customer features into the trained CNN model. Since the CNN model has learned how to identify the person holding the call number paper from the image, it can accurately output the probability of each person being a target customer. If the CNN model predicts that the probability of a person being a target customer exceeds the set threshold, the robot customer service will mark him or her as a target customer to be confirmed.

[0179] Path planning module: used to find a safe and shortest path on an obstacle map.

[0180] Automatic navigation module: The movement of the robot customer service needs to convert the planned path into the movement command of the robot customer service. For example, if the next grid is to the north of the current grid, let the robot customer service move one grid to the north. Send these movement commands to the robot customer service's motion control system (such as motors, sensors, etc.), and the robot customer service will move according to these commands until it reaches the location of the target customer. This module can adjust the movement state of the robot customer service through a closed-loop feedback system to ensure that the robot customer service moves along the path accurately and stably.

[0181] Service interaction module: This module is used for the interaction between the robot customer service and the customer. When the robot customer service arrives at the target customer's location, it needs to ask the person whether he is the target customer through voice or display. If not, it will continue to look for the target customer. If yes, the service interaction module will ask the customer what services he needs through voice or touch screen, and receive the customer's instructions. According to the customer's answer, the robot customer service will query the relevant information in the background service system and provide services to the customer through voice or display.

[0182] Based on the above, a customer touch system suitable for robot customer service is shown in Fig.10 The specific implementation of the customer contact method applicable to robot customer service provided by the present invention includes the following steps:

[0183] S1: The computer vision module collects environmental image data and generates an obstacle map, and then feeds the image data and map into the target recognition module.

[0184] S2: The target recognition module identifies customers who need service and marks them on the map.

[0185] S3: The path planning module plans the optimal path based on the map and target location.

[0186] Specifically, step S3 includes the following contents:

[0187] On the two-dimensional grid map obtained by the computer vision module, the robot customer service and the target customer have their own positions. These two positions can be represented by the coordinates of the grid. For example, the position of the robot customer service may be (x1, y1), and the position of the target customer may be (x2, y2). The cost of each grid consists of two parts: the actual cost from the starting point to this grid, and the predicted cost from this grid to the end point. The actual cost can be obtained by adding the cost of each step of the robot customer service moving from the starting point to this grid (in this scenario, the cost of each step can be considered equal). The predicted cost is usually represented by a linear or quadratic function of the Euclidean distance from this grid to the end point. First, add the starting point to the open list. Then, select the grid with the smallest cost from the open list, check its adjacent grids, and update their costs and parent nodes according to the rules. This process continues until the end point is added to the open list or the open list is empty. Once the target customer is added to the open list, it means that a path has been found. Then starting from the end point, backtracking along the parent node of each grid, a path from the starting point to the end point can be obtained. Then, by reversing this path, we get the path from the starting point to the end point. The path planning module will send the optimal path information to the next level module for the robot customer service to navigate.

[0188] S4: The automatic navigation module guides the robot to move according to this path.

[0189] S5: After arriving at the customer's location, the service interaction module begins to interact with the customer to complete the service process.

[0190] In addition, mobile service robots can interact with backend servers through wireless networks to obtain the latest service information in real time, or upload service records and other data, providing banks with more accurate service management capabilities.

[0191] From the above description, it can be seen that the specific implementation of the present invention provides a customer contact method suitable for robot customer service, including: first, screening out target customers based on the customer's service identifier and a pre-generated service feature extraction model; wherein the target customer is the customer to be served by the robot customer service; then, determining the optimal path between the target customer and the robot customer service based on the target customer's location, the robot customer service location, or the location of customers around the target customer, or the location of customers around the robot customer service; finally, providing customer contact services to the target customer according to the optimal path. The present invention can bring the following main beneficial effects:

[0192] (1) Improve service efficiency: Through the computer vision module and target recognition module, the robot can automatically identify and locate customers who need service, greatly shortening the service response time. In addition, through the path planning module and automatic navigation module, the robot can quickly and accurately move to the customer's location and provide instant service, further improving service efficiency.

[0193] (2) Improve service quality: Robots can provide one-stop services, including consulting, printing and other services, to meet the diverse needs of customers. At the same time, the services provided by robots are program-based, avoiding possible mistakes in manual services and ensuring the quality and accuracy of services.

[0194] (3) Enhanced service accessibility: The robot’s mobility enables it to cover every corner of the bank. No matter where the customer is in the service hall, the robot can proactively come forward to provide services, significantly enhancing the accessibility of banking services.

[0195] (4) Optimizing service experience: Robots can provide multiple services continuously, avoiding the trouble of customers moving back and forth between different service desks, greatly optimizing the customer service experience.

[0196] The above four points together make banking business processing more efficient, more stable in quality, wider in coverage and better in experience.

[0197] Based on the same inventive concept, the embodiments of the present application also provide a customer contact device suitable for robot customer service, which can be used to implement the methods described in the above embodiments, such as the following embodiments. Since the principle of solving the problem by the customer contact device suitable for robot customer service is similar to the customer contact method suitable for robot customer service, the implementation of the customer contact device suitable for robot customer service can refer to the implementation of the customer contact method suitable for robot customer service, and the repeated parts will not be repeated. As used below, the terms "unit" or "module" can be a combination of software and / or hardware that implements predetermined functions. Although the system described in the following embodiments is preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceived.

[0198] The embodiment of the present invention provides a specific implementation of a customer contact device applicable to robot customer service that can implement a customer contact method applicable to robot customer service, see Fig.11 , a customer contact device suitable for robot customer service specifically includes the following contents:

[0199] The target customer screening module 10 is used to screen out target customers according to the customer's service identification and the pre-generated service feature extraction model; wherein the target customer is the customer to be served by the robot customer service;

[0200] An optimal path determination module 20, configured to determine an optimal path between the target customer and the robot customer service according to the location of the target customer, the location of the robot customer service, and at least one of the locations of customers around the target customer and the locations of customers around the robot customer service;

[0201] The customer contact service module 30 is used to provide customer contact service to the target customer according to the optimal path.

[0202] In some embodiments of the present invention, see Fig.12 , a customer contact device suitable for robot customer service, further comprising:

[0203] Model generation module 40, used to generate the service feature extraction model, see Fig.13 , the model generation module 40 includes:

[0204] An image data acquisition unit 40a, configured to acquire image data corresponding to the historical data of the service identifier;

[0205] The model generation unit 40b is used to train the initial model of the service feature extraction model according to the pixel values ​​of the image data to generate the service feature extraction model; wherein the initial model is a convolutional recurrent neural network.

[0206] In some embodiments of the present invention, see Fig.14 , the model generation unit 40b includes:

[0207] A feature extraction unit 40b1, used for inputting the pixel value into the convolution layer in the initial model to extract the features of the service identifier;

[0208] A sequence data extraction unit 40b2, used for inputting the features of the service identifier into the recurrent layer in the initial model to extract the sequence data of the service identifier; wherein the sequence data is used to characterize the relationship between the features of the service identifier and its context;

[0209] A character distribution generating unit 40b3, used for inputting the features and the sequence data into the connection layer of the initial model to generate a character probability distribution corresponding to the image data;

[0210] The model training unit 40b4 is used to perform forward propagation and / or backward propagation on the initial model according to the character probability distribution and a preset threshold.

[0211] In some embodiments of the present invention, see Fig.15 , the target customer screening module 10 comprises:

[0212] A grayscale image generating unit 10a, configured to convert the image data corresponding to the service identifier into a grayscale image;

[0213] A threshold processing result generating unit 10b, configured to perform threshold processing on the grayscale image to generate a threshold processing result;

[0214] A probability distribution generating unit 10c, used for inputting the threshold processing result into the service feature extraction model to generate the character probability distribution;

[0215] A text sequence determination unit 10d, configured to determine a text sequence corresponding to the service identifier according to the character probability distribution;

[0216] The target customer screening unit 10e is used to screen out the target customers according to the text sequence and the local customer contact service scope.

[0217] In some embodiments of the present invention, see Fig.16 , the service identifier is the queue number of the service that the customer wants to obtain; the customer contact device also includes:

[0218] The real-time location monitoring module 50 is used to monitor the location of the customer in real time after the customer obtains the queue number and the service the customer wants to obtain belongs to the local contact service range.

[0219] In some embodiments of the present invention, see Fig.17 When there are no customers around the robot customer service and there are no customers around the target customer, the optimal path determination module 20 includes:

[0220] A grid map generating unit 20a, configured to generate a grid map corresponding to the scene where the robot customer service and the target customer are located;

[0221] The mobile cost determination unit 20b is used to determine the mobile cost of each grid in the grid map according to the location of the target customer and the location of the robot customer service; wherein the mobile cost includes an actual cost and a predicted cost; the actual cost is the path loss of the robot customer service moving to the current grid, and the predicted cost is the path loss of the robot moving from the current grid to the location of the target customer;

[0222] The optimal path determining unit 20c is used to determine the optimal path according to the movement cost of each grid.

[0223] In some embodiments of the present invention, see Fig.18 When there are customers around the robot customer service and / or there are customers around the target customer, the optimal path determination module 20 includes:

[0224] An initial path generating unit 20d, configured to generate an initial path according to the movement cost of each grid;

[0225] An alternative path generating unit 20e, configured to generate at least one alternative path according to the initial path and in combination with at least one of the positions of customers around the target customer and the positions of customers around the robot customer service;

[0226] The path adjustment unit 20f is used to adjust the path in the initial path and the at least one alternative path in real time according to the current locations of the customers around the target customer and / or the current locations of the customers around the robot customer service to determine the optimal path.

[0227] From the above description, it can be seen that an embodiment of the present invention provides a customer contact device suitable for robot customer service, including: a target customer screening module, used to screen out target customers according to the customer's service identification and a pre-generated service feature extraction model; wherein the target customer is a customer to whom the robot customer service is to provide service; an optimal path determination module, used to determine the optimal path between the target customer and the robot customer service according to the location of the target customer, the location of the robot customer service, or the location of customers around the target customer, or the location of customers around the robot customer service; a customer contact service module, used to provide customer contact service to the target customer according to the optimal path.

[0228] The present invention provides a touch device for robot customer service, which enables the robot customer service to automatically move to the location of the customer who needs service and provide one-stop service, thereby greatly improving service efficiency and enhancing service experience. At the same time, the mobility of the robot customer service also solves the accessibility problem of fixed self-service equipment. In addition, the robot customer service can provide a series of services from simple to complex, making the service experience more coherent and reducing the waiting time for customers. This mobile service robot customer service provides a new service mode for the service industry, which can further improve service efficiency and customer satisfaction.

[0229] It should be noted that the customer contact method and device for robot customer service provided in the embodiments of the present invention can be used in the financial field, and can also be used in any technical field other than the financial field. The embodiments of the present invention do not limit the application field of the customer contact method and device for robot customer service.

[0230] The embodiments of the present application also provide a specific implementation of an electronic device that can implement all the steps in the customer contact method applicable to robot customer service in the above embodiments, see Fig.19 , electronic equipment specifically includes the following:

[0231] Processor processor) 1201, memory memory) 1202, communication interface Communications Interface) 1203 and bus 1204;

[0232] The processor 1201, the memory 1202, and the communication interface 1203 communicate with each other through the bus 1204; the communication interface 1203 is used to realize information transmission between the server-side device and the user-side device and other related devices;

[0233] The processor 1201 is used to call the computer program in the memory 1202. When the processor executes the computer program, all the steps in the customer contact method applicable to the robot customer service in the above embodiment are implemented. For example, when the processor executes the computer program, the following steps are implemented:

[0234] Step 100: Screening out target customers according to the customer's service identifier and the pre-generated service feature extraction model; wherein the target customer is the customer to be served by the robot customer service;

[0235] Step 200: Determine an optimal path between the target customer and the robot customer service according to the location of the target customer, the location of the robot customer service, and at least one of the locations of customers around the target customer and the locations of customers around the robot customer service;

[0236] Step 300: Provide customer service to the target customer according to the optimal path.

[0237] The embodiments of the present application also provide a computer-readable storage medium capable of implementing all the steps of the customer contact method applicable to robot customer service in the above embodiments. A computer program is stored on the computer-readable storage medium. When the computer program is executed by a processor, all the steps of the customer contact method applicable to robot customer service in the above embodiments are implemented. For example, when the processor executes the computer program, the following steps are implemented:

[0238] Step 100: Screening out target customers according to the customer's service identifier and the pre-generated service feature extraction model; wherein the target customer is the customer to be served by the robot customer service;

[0239] Step 200: Determine an optimal path between the target customer and the robot customer service according to the location of the target customer, the location of the robot customer service, and at least one of the locations of customers around the target customer and the locations of customers around the robot customer service;

[0240] Step 300: Provide customer service to the target customer according to the optimal path.

[0241] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the hardware + program embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0242] The above is a description of a specific embodiment of the specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0243] Although the present application provides method operation steps such as embodiments or flow charts, more or fewer operation steps may be included based on conventional or non-creative labor. The order of steps listed in the embodiments is only one way of executing the order of many steps and does not represent the only execution order. When the device or user-end product in practice is executed, it can be executed in the order of the method shown in the embodiments or the drawings or in parallel (for example, in a parallel processor or multi-threaded processing environment).

[0244] For the convenience of description, the above devices are described in various modules according to their functions. Of course, when implementing the embodiments of this specification, the functions of each module can be implemented in the same or more software and / or hardware, or the module implementing the same function can be implemented by a combination of multiple sub-modules or sub-units. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0245] Those skilled in the art also know that, in addition to implementing the controller in a purely computer-readable program code, the controller can be made to implement the same function in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers by logically programming the method steps. Therefore, such a controller can be considered as a hardware component, and the devices for implementing various functions included therein can also be considered as structures within the hardware component. Or even, the devices for implementing various functions can be considered as both software modules for implementing the method and structures within the hardware component.

[0246] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0247] The memory may include non-permanent storage in a computer-readable medium, in the form of random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash memory (RAM). The memory is an example of a computer-readable medium.

[0248] Each embodiment in this specification is described in a progressive manner, and the same and similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment. In the description of this specification, the description of the reference terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the embodiment of this specification. In this specification, the schematic representation of the above terms does not necessarily target the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner. In addition, in the absence of contradiction, a person skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of different embodiments or examples.

[0249] The above is only an example of the embodiment of the present specification and is not intended to limit the embodiment of the present specification. For those skilled in the art, the embodiment of the present specification may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the embodiment of the present specification shall be included in the scope of the claims of the embodiment of the present specification.

Claims

1. A customer contact method suitable for robot customer service, characterized in that: include: Filter out target customers according to the customer's service identifier and the pre-generated service feature extraction model; wherein the target customers are customers to be served by the robot customer service; Determine an optimal path between the target customer and the robot customer service according to the location of the target customer, the location of the robot customer service, and at least one of the locations of customers around the target customer and the locations of customers around the robot customer service; Provide customer service to the target customer according to the optimal path.

2. The customer contact method according to claim 1, characterized in that: The steps of generating the service feature extraction model include: Acquire image data corresponding to the historical data of the service identifier; The initial model of the service feature extraction model is trained according to the pixel values ​​of the image data to generate the service feature extraction model; wherein the initial model is a convolutional recurrent neural network.

3. The customer contact method according to claim 2, characterized in that: Training the initial model of the service feature extraction model according to the pixel values ​​of the image data includes: Inputting the pixel value into a convolutional layer in the initial model to extract features of the service identifier; Inputting the features of the service identifier into the recurrent layer in the initial model to extract sequence data of the service identifier; wherein the sequence data is used to characterize the relationship between the features of the service identifier and its context; Inputting the features and the sequence data into the connection layer of the initial model to generate a character probability distribution corresponding to the image data; The initial model is forward propagated and / or backward propagated according to the character probability distribution and a preset threshold.

4. The customer contact method according to claim 3, characterized in that: The step of screening out target customers according to the customer's service identifier and the pre-generated service feature extraction model includes: Converting the image data corresponding to the service identifier into a grayscale image; Performing threshold processing on the grayscale image to generate a threshold processing result; Inputting the threshold processing result into the service feature extraction model to generate the character probability distribution; Determining a text sequence corresponding to the service identifier according to the character probability distribution; The target customers are screened out based on the text sequence and the local customer contact service scope.

5. The customer contact method according to claim 4, characterized in that: The service identifier is the queue number of the service that the customer wants to obtain; the customer contact method also includes: After the customer obtains the queue number and the service the customer wants to obtain falls within the scope of the local contact service, the customer's location is detected.

6. The customer contact method according to any one of claims 1 to 5, characterized in that: When there are no customers around the robot customer service and there are no customers around the target customer, determining an optimal path between the target customer and the robot customer service according to the location of the target customer and the location of the robot customer service, including: Generate a grid map corresponding to the scene where the robot customer service and the target customer are located; Determine the movement cost of each grid in the grid map according to the location of the target customer and the location of the robot customer service; wherein the movement cost includes an actual cost and a predicted cost; the actual cost is the path loss of the robot customer service moving to the current grid, and the predicted cost is the path loss of the robot moving from the current grid to the location of the target customer; An optimal path is determined according to the movement cost of each grid.

7. The customer contact method according to claim 6, characterized in that: When there are customers around the robot customer service and / or there are customers around the target customer, determining the optimal path between the target customer and the robot customer service according to the location of the target customer, the location of the robot customer service, and combining at least one of the locations of customers around the target customer and the locations of customers around the robot customer service, including: Generate an initial path according to the movement cost of each grid; Generate at least one alternative path based on the initial path and in combination with at least one of the locations of customers around the target customer and the locations of customers around the robot customer service; The path is adjusted in real time in the initial path and the at least one alternative path according to the current locations of the customers around the target customer and / or the current locations of the customers around the robot customer service to determine the optimal path.

8. A customer contact device suitable for robot customer service, characterized in that: include: A target customer screening module is used to screen out target customers according to the customer's service identification and a pre-generated service feature extraction model; wherein the target customer is a customer to be served by the robot customer service; an optimal path determination module, configured to determine an optimal path between the target customer and the robot customer service according to the location of the target customer, the location of the robot customer service, and at least one of the locations of customers around the target customer and the locations of customers around the robot customer service; The customer contact service module is used to provide customer contact service to the target customer according to the optimal path.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the steps of the customer contact method applicable to robot customer service as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the customer contact method applicable to robot customer service as described in any one of claims 1 to 7 are implemented.

Images