Object incoming line quantity analysis method and device, electronic equipment, medium and product

Abstract

The embodiment of the invention discloses an object incoming line quantity analysis method and device, electronic equipment, a storage medium and a product, and the method comprises the steps: obtaining a plurality of to-be-analyzed feature data of each target object set in a current incoming line quantity analysis period, the incoming line quantity of the first object is associated with at least one second object; inputting the plurality of to-be-analyzed feature data into a pre-trained joint incoming line quantity analysis and prediction model to obtain a plurality of groups of first object incoming line quantity prediction results of each target object set; and based on the actual incoming line quantity data of each target object set in at least one adjacent historical incoming line quantity analysis period, performing verification analysis on the corresponding multiple groups of first object incoming line quantity prediction results, and determining a corresponding target first object incoming line quantity prediction result. According to the technical scheme provided by the embodiment of the invention, the second-line customer service incoming line quantity can be predicted by integrating the multi-dimensional feature data influencing the incoming line quantity, and the prediction deviation is reduced.

Description

Technical Field

[0001] The embodiments of the present disclosure relate to the field of artificial intelligence technology, and in particular to a method, device, electronic device, medium and product for analyzing incoming line quantity of an object. Background Art

[0002] In the commercial service sector, business volumes are usually estimated as reference data for making staff schedules, in the hope of meeting customer service needs while minimizing labor redundancy and reducing costs.

[0003] Customer service includes two levels: first-line customer service and second-line customer service. Second-line customer service usually needs to handle customer problems that first-line customer service cannot solve, and provide customers with more professional services that can meet customer needs.

[0004] However, since the business volume of second-line customer service is affected by many factors such as the overall business volume, customer service skill grouping, and the service quality of first-line customer service, and the business chain is long, the prediction results will have large deviations if only the prediction analysis is based on historical business volume data. Summary of the invention

[0005] The present disclosure provides an object incoming line quantity analysis method, device, electronic device, medium and product, which can integrate multi-dimensional characteristic data that affect the incoming line quantity, predict the incoming line quantity of a target object whose incoming line quantity is affected by other related objects, reduce the prediction deviation, and thereby enable tasks based on the prediction results to obtain better task effects.

[0006] In a first aspect, an embodiment of the present disclosure provides a method for analyzing incoming line quantity of an object, the method comprising:

[0007] Acquire multiple pieces of feature data to be analyzed for each target object set in a current incoming line quantity analysis cycle, wherein the target object set includes at least one first object, and the incoming line quantity of the first object is associated with at least one second object;

[0008] Inputting the plurality of feature data to be analyzed into a pre-trained joint incoming line quantity analysis and prediction model to obtain a plurality of groups of first object incoming line quantity prediction results for each target object set;

[0009] Based on the actual incoming line quantity data of each target object set in at least one adjacent historical incoming line quantity analysis cycle, the corresponding multiple groups of first object incoming line quantity prediction results are verified and analyzed to determine the corresponding target first object incoming line quantity prediction results.

[0010] In a second aspect, the present disclosure also provides a device for analyzing incoming line quantity of an object, the device comprising:

[0011] A feature data acquisition unit, used to acquire multiple feature data to be analyzed for each target object set in a current incoming line quantity analysis cycle, wherein the target object set includes at least one first object, and the incoming line quantity of the first object is associated with at least one second object;

[0012] An incoming line quantity prediction unit, used for inputting the plurality of feature data to be analyzed into a pre-trained joint incoming line quantity analysis prediction model to obtain a plurality of groups of first object incoming line quantity prediction results for each target object set;

[0013] The incoming line quantity prediction result determination unit is used to verify and analyze the corresponding multiple groups of the first object incoming line quantity prediction results based on the actual incoming line quantity data of each target object set in at least one adjacent historical incoming line quantity analysis cycle, and determine the corresponding target first object incoming line quantity prediction result.

[0014] In a third aspect, an embodiment of the present disclosure further provides an electronic device, the electronic device comprising:

[0015] one or more processors;

[0016] a storage device for storing one or more programs,

[0017] When the one or more programs are executed by the one or more processors, the one or more processors implement the object incoming line quantity analysis method as described in any of the embodiments of the present disclosure.

[0018] In a fourth aspect, the embodiments of the present disclosure further provide a storage medium comprising computer executable instructions, which, when executed by a computer processor, are used to execute the object incoming line quantity analysis method as described in any of the embodiments of the present disclosure.

[0019] In a fifth aspect, the embodiments of the present disclosure further provide a computer program product, including a computer program, which, when executed by a processor, implements the object incoming line quantity analysis method as described in any one of the embodiments of the present invention.

[0020] In the disclosed embodiment, multiple feature data to be analyzed are obtained for each target object set in the current incoming line analysis cycle, wherein the target object set includes at least one first object, and the incoming line volume of the first object has an associated relationship with at least one second object; the multiple feature data to be analyzed are input into a pre-trained joint incoming line analysis prediction model to obtain multiple groups of first object incoming line prediction results for each target object set; based on the actual incoming line data of each target object set in at least one adjacent historical incoming line analysis cycle, the corresponding multiple groups of first object incoming line prediction results are verified and analyzed to determine the corresponding target first object incoming line prediction results. The technical solution of the disclosed embodiment can be applied to the second-line customer service incoming line volume prediction scenario, which solves the problem of large deviation in the current prediction of the incoming line volume for the second-line customer service with a longer business link length, and can comprehensively predict the incoming line volume of the second-line customer service with multi-dimensional feature data that affects the incoming line volume, thereby reducing the prediction deviation, and further combines the historical incoming line volume data to determine the optimal prediction result among multiple groups of prediction results. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The above and other features, advantages and aspects of the embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the accompanying drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and the originals and elements are not necessarily drawn to scale.

[0022] Figure 1 It is a flow chart of a method for analyzing incoming line quantity of an object provided by an embodiment of the present disclosure;

[0023] Figure 2 It is a flow chart of a method for analyzing incoming line quantity of an object provided by an embodiment of the present disclosure;

[0024] Figure 3 It is a flow chart of a method for analyzing incoming line quantity of an object provided by an embodiment of the present disclosure;

[0025] Figure 4 It is a structural schematic diagram of an object incoming line quantity analysis device provided by an embodiment of the present disclosure;

[0026] Figure 5 It is a structural schematic diagram of an electronic device provided by an embodiment of the present disclosure. DETAILED DESCRIPTION

[0027] Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as being limited to the embodiments described herein, which are instead provided for a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes and are not intended to limit the scope of protection of the present disclosure.

[0028] It should be understood that the various steps described in the method embodiments of the present disclosure may be performed in different orders and / or in parallel. In addition, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present disclosure is not limited in this respect.

[0029] The term "including" and its variations used herein are open inclusions, i.e., "including but not limited to". The term "based on" means "based at least in part on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". The relevant definitions of other terms will be given in the following description.

[0030] It should be noted that the concepts such as "first" and "second" mentioned in the present disclosure are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.

[0031] It should be noted that the modifications of "one" and "plurality" mentioned in the present disclosure are illustrative rather than restrictive, and those skilled in the art should understand that unless otherwise clearly indicated in the context, it should be understood as "one or more".

[0032] It is understandable that before using the technical solutions disclosed in the embodiments of the present disclosure, the types, scope of use, usage scenarios, etc. of the personal information involved in the present disclosure should be informed to the user and the user's authorization should be obtained in an appropriate manner in accordance with relevant laws and regulations.

[0033] For example, in response to receiving an active request from a user, a prompt message is sent to the user to clearly prompt the user that the operation requested to be performed will require obtaining and using the user's personal information. Thus, the user can autonomously choose whether to provide personal information to software or hardware such as an electronic device, application, server, or storage medium that performs the operation of the technical solution of the present disclosure according to the prompt message.

[0034] As an optional but non-limiting implementation, in response to receiving an active request from the user, the prompt information may be sent to the user in the form of a pop-up window, in which the prompt information may be presented in text form. In addition, the pop-up window may also carry a selection control for the user to choose "agree" or "disagree" to provide personal information to the electronic device.

[0035] It is understandable that the above notification and the process of obtaining user authorization are merely illustrative and do not constitute a limitation on the implementation of the present disclosure. Other methods that meet the relevant laws and regulations may also be applied to the implementation of the present disclosure.

[0036] It is understandable that the data involved in this technical solution (including but not limited to the data itself, the acquisition or use of the data) shall comply with the requirements of relevant laws, regulations and relevant provisions.

[0037] Figure 1 A flow chart of an object incoming call volume analysis method provided in an embodiment of the present disclosure is applicable to scenarios in which the incoming call volume of a target object whose incoming call volume is affected by other objects is predicted, especially in scenarios in which predictive analysis is performed on customer service incoming call volume, and accurate prediction is made on the incoming call volume of second-line customer service with longer business links. The method can be executed by an object incoming call volume analysis device, which can be implemented in the form of software and / or hardware. Optionally, the object incoming call volume analysis device can be implemented by an electronic device, which can be a mobile terminal, a PC or a server, etc.

[0038] like Figure 1 As shown, the object incoming line quantity analysis method includes:

[0039] S110, obtaining multiple pieces of feature data to be analyzed for each target object set in a current incoming line quantity analysis cycle, wherein the target object set includes at least one first object, and the incoming line quantity of the first object is associated with at least one second object.

[0040] In the customer service business, a first object and a second object are usually included. Among them, the first object is, for example, the second-line customer service, and the second object is, for example, the first-line customer service. There is a correlation between the incoming call volume of the first object and the second object. The first-line customer service is usually the customer service representative who first contacts the customer. He is responsible for handling common and simple problems, providing basic after-sales support and answering common questions. When customers encounter more complex, technical or professional problems, the first-line customer service may not be able to provide a satisfactory solution. At this time, the customer service work order will be sent to the second-line customer service for processing. The second-line customer service usually works closely with the first-line customer service to help solve more complex problems and ensure that customers receive comprehensive support and satisfactory solutions. Therefore, the incoming call volume of the first object is affected by the incoming call volume of the second object.

[0041] In this business scenario, the target object set is, for example, a second-line customer service work group. A second-line customer service work group refers to a work management group for second-line customer service staff, which can be divided according to different skills or professional fields. Each second-line customer service work group is composed of customer service staff with similar skills and knowledge reserves, so as to more effectively handle customer service tasks such as specific types of calls, inquiries or problems.

[0042] Since the service content and business volume of each second-line customer service working group have different business characteristics, in order to ensure the accuracy of the incoming call volume prediction results of each second-line customer service working group, in this embodiment, the incoming call volume of each second-line customer service working group is predicted and analyzed separately. Among them, the incoming call volume can be understood as the number of calls or consultations received by the customer service center or customer service team within a certain period of time. It represents the amount of interaction between customers and the customer service team, and can be used to evaluate the workload, resource requirements and operational efficiency of the customer service team. The forecast analysis of the incoming call volume of each second-line customer service working group is the number of service businesses such as calls or consultations that each second-line customer service working group may receive in the future.

[0043] The current incoming line analysis cycle can be understood as the time interval for which incoming line forecasting is required. For example, if a preset number of days (such as 10 days) is required in advance to forecast incoming line for a week, then the current incoming line analysis cycle is the week after the preset number of days from the current time.

[0044] The feature data to be analyzed can be any data that has an impact on the forecast result of the incoming call volume analysis. For example, feature data such as historical incoming call volume data, holidays, sales activity events, first-line customer service business processing data, and service complaint rate. Different second-line customer service working groups can correspond to different feature data items to be analyzed. In this embodiment, different multiple feature data to be analyzed can be selected for each different second-line customer service working group as reference data for incoming call volume forecasting. For example, the degree of influence of each feature data item on the incoming call volume forecast result can be determined in combination with the skill characteristics of each second-line customer service working group, and multiple feature data to be analyzed with a higher degree of influence on the incoming call volume forecast result can be selected for each second-line customer service working group.

[0045] S120, inputting the plurality of feature data to be analyzed into a pre-trained joint incoming line quantity analysis and prediction model to obtain a plurality of groups of first object incoming line quantity prediction results for each of the target object sets.

[0046] The joint incoming line quantity analysis and prediction model is a trained neural network model with the function of incoming line quantity prediction. The joint incoming line quantity analysis and prediction model includes multiple independent incoming line quantity analysis and prediction network modules and a prediction result fusion module. Therefore, the model can perform feature analysis on multiple input feature data to be analyzed and output multiple groups of first object incoming line quantity prediction results.

[0047] In an optional embodiment, the independent incoming line analysis prediction network module may include at least two model structures of an autoregressive exponentially weighted average moving model, a time series model, an extreme gradient boosting multi-factor model (machine learning model), and a deep learning model with a preset structure. Through model structures with different data analysis / learning features / capabilities, multiple feature data to be analyzed can be analyzed from different data analysis dimensions to reduce the prediction error. For example, the time series model can extract the temporal features of each feature data to be analyzed, and the extreme gradient boosting multi-factor model can comprehensively analyze the correlation features of multiple feature data to be analyzed.

[0048] Based on the model structure of the joint incoming line quantity analysis and prediction model, obtaining multiple groups of first object incoming line quantity prediction results for each target object set may be performed as follows:

[0049] First, multiple feature data to be analyzed are respectively input into each independent incoming line quantity analysis and prediction module in the joint incoming line quantity analysis and prediction model to obtain multiple sets of first object incoming line quantity independent prediction results. For example, the exponentially weighted moving average model (Exponentially Weighted Moving-Average, EWMA), prophet model, XGB model (eXtreme Gradient Boosting) and the deep learning model with a preset structure respectively perform data analysis on multiple feature data to be analyzed and output incoming line quantity predictions, and correspondingly 4 sets of first object incoming line quantity independent prediction results can be obtained.

[0050] Then, the four groups of independent prediction results of the first object incoming line quantity obtained in the previous step can be input into the prediction result fusion module in the joint incoming line quantity analysis and prediction model to obtain the corresponding first object incoming line quantity fusion prediction result.

[0051] In summary, based on the joint incoming line quantity analysis and prediction model, five groups of first object incoming line quantity fusion prediction results can be obtained. These five groups of first object incoming line quantity fusion prediction results can be used as candidate values ​​for the final prediction result.

[0052] S130. Based on the actual incoming line volume data of each target object set in at least one adjacent historical incoming line volume analysis period, verify and analyze the corresponding multiple groups of first object incoming line volume prediction results to determine the corresponding target first object incoming line volume prediction results.

[0053] Based on the five groups of first object incoming line quantity fusion prediction results obtained in the previous step, when determining the optimal prediction result, it is not practical to solve it directly according to the integer programming problem. This is because the problem scale is large and it is difficult to directly model and solve it; moreover, the optimization goal is to maximize the prediction accuracy of the future time, but the prediction accuracy of the future is something that has not happened and the corresponding data cannot be obtained. Therefore, in this embodiment, the accuracy maximization between the actual value of at least one historical incoming line quantity analysis cycle and the predicted value output by the current incoming line quantity analysis cycle is used as an approximate goal to determine the final target first object incoming line quantity prediction result.

[0054] Specifically, the prediction deviation value between each group of first object incoming line quantity prediction results and the incoming line quantity data corresponding to each time node in the actual incoming line quantity data can be calculated respectively. Among them, the time node refers to multiple time nodes in an incoming line quantity analysis cycle. For example, an incoming line quantity analysis cycle is a week, including seven days, and correspondingly, it includes 7 time nodes from Monday, Tuesday to Sunday. The prediction results of each time node in each group of first object incoming line quantity prediction results can be used to calculate the prediction deviation value of the actual incoming line quantity at the same time point in at least one adjacent incoming line quantity analysis cycle. The prediction deviation value can be an indicator for measuring the prediction accuracy. The corresponding prediction deviation value is determined by calculating the ratio of the difference between the incoming line quantity prediction value and the actual incoming line quantity value to the corresponding incoming line quantity actual value.

[0055] Then, based on the prediction deviation value and the preset prediction compliance deviation standard, the compliance rate of each group of the first object incoming line quantity prediction results is determined; and a group of the first object incoming line quantity prediction results corresponding to the maximum compliance rate value is used as the corresponding target first object incoming line quantity prediction results.

[0056] In a specific example, the calculation process can be expressed as follows: [ti1, ti2, ti3, ti4, ti5, ti6, ti7] represents the forecast result of the incoming call volume of the second-line customer service of the i-th group. [rj1, rj2, rj3, rj4, rj5, rj6, rj7] represents the actual incoming call volume data in the adjacent j historical incoming call volume analysis cycles. Considering that the incoming call volume of the second-line customer service has a strong trend during the week and weekend, the forecast compliance ratio will be calculated separately during the week and weekend. The forecast deviations from Monday to Sunday are: (ti1-rj1) / rj1, (ti1-rj2) / rj2, (ti1-rj3) / rj3, (ti1-rj4) / rj4, (ti1-rj5) / rj5, (ti6-rj6) / rj6, (ti7-rj7) / rj7.

[0057] Among them, the value of j is 1 to 4 (representing 4 weeks of history). By analogy, the forecast deviations from Monday to Friday, Saturday and Sunday can be calculated. Among them, the preset forecast compliance deviation standard can be considered to be in the range of [-10%, 8%]. Then the compliance rate of this group of forecast values ​​in the past 4 weeks is the number of forecast deviations from Monday to Friday within the forecast compliance deviation standard range / (5*4), and the number of forecast deviations on Saturday and Sunday within the forecast compliance deviation standard range / (2*4). According to this calculation process, the forecast compliance rate of each group of second-line customer service incoming call forecast results can be calculated, and the group of second-line customer service incoming call forecast results with the largest forecast compliance rate can be selected as the target second-line customer service incoming call forecast result.

[0058] The technical solution of the disclosed embodiment is to obtain multiple feature data to be analyzed for each target object set in the current incoming line analysis cycle, wherein the target object set includes at least one first object, and the incoming line volume of the first object has an associated relationship with at least one second object; the multiple feature data to be analyzed are input into a pre-trained joint incoming line analysis prediction model to obtain multiple groups of first object incoming line prediction results for each target object set; based on the actual incoming line data of each target object set in at least one adjacent historical incoming line analysis cycle, the corresponding multiple groups of first object incoming line prediction results are verified and analyzed to determine the corresponding target first object incoming line prediction results. The technical solution of the disclosed embodiment solves the problem of large deviation in the current prediction of the incoming line volume for second-line customer service with longer service links, and can predict the incoming line volume of second-line customer service by integrating multi-dimensional feature data that affect the incoming line volume, thereby reducing the prediction deviation, and further combines the historical incoming line volume data to determine the optimal prediction result among multiple groups of prediction results.

[0059] Figure 2 This is a flow chart of another object incoming call analysis method provided by the embodiment of the present disclosure. On the basis of the above embodiment, it further explains the process of obtaining multiple groups of first object (second-line customer service) incoming call prediction results. The method can be executed by an object incoming call analysis device, which can be implemented in the form of software and / or hardware, and optionally, by an electronic device, which can be a mobile terminal, a PC or a server.

[0060] like Figure 2 As shown, the object incoming line quantity analysis method includes:

[0061] S210. Obtain multiple pieces of feature data to be analyzed for each second-line customer service working group in the current incoming call volume analysis cycle.

[0062] S220. Input the plurality of feature data to be analyzed into each independent incoming call volume analysis and prediction module in the joint incoming call volume analysis and prediction model respectively to obtain a plurality of groups of independent prediction results of the second-line customer service incoming call volume.

[0063] The joint incoming call volume analysis and prediction model is a trained neural network model with the function of incoming call volume prediction. The joint incoming call volume analysis and prediction model includes multiple independent incoming call volume analysis and prediction network modules and a prediction result fusion module. Therefore, the model can perform feature analysis on multiple input feature data to be analyzed and output multiple sets of second-line customer service incoming call volume prediction results.

[0064] The independent incoming line analysis and prediction network module may include at least two model structures among an autoregressive exponentially weighted average moving model, a time series model, an extreme gradient boosting multi-factor model (machine learning model), and a deep learning model with a preset structure. Through model structures with different data analysis / learning features / capabilities, multiple feature data to be analyzed can be analyzed from different data analysis dimensions to reduce the prediction error. For example, the time series model can extract the temporal features of each feature data to be analyzed, and the extreme gradient boosting multi-factor model can comprehensively analyze the correlation features of multiple feature data to be analyzed.

[0065] Multiple feature data to be analyzed can be input into each independent incoming line analysis and prediction module in the joint incoming line analysis and prediction model to obtain multiple sets of independent prediction results for the incoming line volume of second-line customer service. For example, the exponentially weighted moving average model (Exponentially Weighted Moving-Average, EWMA), prophet model, XGB model (eXtreme Gradient Boosting) and deep learning model with preset structure respectively analyze multiple feature data to be analyzed and output incoming line volume prediction, and accordingly, 4 sets of independent prediction results for the incoming line volume of second-line customer service can be obtained.

[0066] S230. Input the plurality of groups of independent prediction results of the second-line customer service call volume into the prediction result fusion module in the joint call volume analysis and prediction model to obtain the corresponding second-line customer service call volume fusion prediction result.

[0067] The four groups of independent prediction results of the second-line customer service call volume obtained in the previous step can be input into the prediction result fusion module in the joint call volume analysis and prediction model to obtain the corresponding second-line customer service call volume fusion prediction results.

[0068] The prediction result fusion module is a functional module set up to improve the accuracy of the prediction task results through the joint decision-making of the above-mentioned multiple models. The principle of fusing each group of prediction results is to select models with the best effect and the largest difference as possible for fusion. Among them, the fusion parameters in the prediction result fusion module can also be a set of parameters determined through the model training process that can make the evaluation results of the accuracy of the prediction results perform well.

[0069] S240, performing data enhancement processing on the incoming call prediction results according to the multiple groups of independent prediction results of the second-line customer service incoming call volume and / or the second-line customer service incoming call volume fusion prediction results, to obtain multiple groups of second-line customer service incoming call volume expansion prediction results.

[0070] Furthermore, in order to enrich the candidate prediction result set for determining the target second-line customer service call volume prediction result, in this embodiment, prediction data amplification is performed based on multiple sets of second-line customer service call volume independent prediction results and / or the second-line customer service call volume fusion prediction results.

[0071] Prediction data expansion can be to adjust the values ​​of multiple sets of independent prediction results of second-line customer service call volume and the fusion prediction results of second-line customer service call volume. The adjustment strategy can be to increase or decrease the prediction result value according to the preset adjustment percentage, and can also perform weighted calculation on any combination of multiple sets of independent prediction results of second-line customer service call volume and fusion prediction results of second-line customer service call volume to obtain a new second-line customer service call volume expansion prediction result. Based on a richer set of candidate prediction results, the prediction accuracy can be further improved to a certain extent, which means that the corresponding second-line customer service working group has a stronger tolerance for volatility risks.

[0072] S250. Based on the actual incoming call volume data of each second-line customer service working group in at least one adjacent historical incoming call volume analysis cycle, the corresponding multiple groups of second-line customer service independent customer service incoming call volume prediction results, the second-line customer service incoming call volume fusion prediction results and the multiple groups of second-line customer service incoming call volume amplification prediction results are verified and analyzed respectively to determine the corresponding target second-line customer service incoming call volume prediction results.

[0073] In this step, the process of verifying and analyzing the prediction results of multiple groups of second-line customer service calls in the above embodiment can also be used. For each second-line customer service working group, the corresponding target second-line customer service call prediction results can be determined from the corresponding multiple groups of second-line customer service call independent prediction results, second-line customer service call fusion prediction results and multiple groups of second-line customer service call amplification prediction results.

[0074] The technical solution of the disclosed embodiment is as follows: by obtaining multiple feature data to be analyzed of each second-line customer service working group in the current incoming call volume analysis cycle; inputting the multiple feature data to be analyzed into each independent incoming call volume analysis and prediction module in the joint incoming call volume analysis and prediction model respectively, to obtain multiple groups of second-line customer service independent customer service incoming call volume prediction results; inputting the multiple groups of second-line customer service independent customer service incoming call volume prediction results into the prediction result fusion module in the joint incoming call volume analysis and prediction model to obtain corresponding second-line customer service incoming call volume fusion prediction results; performing data enhancement processing on the incoming call volume prediction result according to the multiple groups of second-line customer service independent customer service incoming call volume prediction results and / or the second-line customer service incoming call volume fusion prediction results to obtain multiple groups of second-line customer service incoming call volume expansion prediction results; based on the actual incoming call volume data of each second-line customer service working group in at least one adjacent historical incoming call volume analysis cycle, respectively verifying and analyzing the corresponding multiple groups of second-line customer service independent customer service incoming call volume prediction results, the second-line customer service incoming call volume fusion prediction results and the multiple groups of second-line customer service incoming call volume expansion prediction results to determine the corresponding target second-line customer service incoming call volume prediction results. The technical solution of the disclosed embodiment solves the problem of large deviation in the current prediction of the call volume of second-line customer service with longer business links. It can predict the call volume of second-line customer service by integrating multi-dimensional characteristic data affecting the call volume, thereby reducing the prediction deviation, and further combines historical call volume data to determine the optimal prediction result from multiple groups of prediction results.

[0075] Figure 3 The present invention provides a flow chart of an object incoming line analysis method, which further explains the training process of the joint incoming line analysis prediction model based on the above embodiment. The method can be performed by an object incoming line analysis device, which can be implemented in the form of software and / or hardware, and optionally, by an electronic device, which can be a mobile terminal, a PC or a server.

[0076] like Figure 3 As shown, the object incoming line quantity analysis method includes:

[0077] S310: Perform model training based on a preset loss function including prediction deviation loss and prediction result compliance rate loss to obtain a target joint incoming line quantity analysis prediction model.

[0078] Among them, the target joint incoming line quantity analysis prediction model includes multiple independent incoming line quantity analysis prediction network modules and a prediction result fusion module. The process of model training based on a preset loss function including prediction deviation loss and prediction result compliance rate loss can be to train the initial joint incoming line quantity analysis prediction model corresponding to the target joint incoming line quantity analysis prediction model through the preset model training samples until the prediction results output by each independent incoming line quantity analysis prediction network module and the loss between the prediction results output by the prediction result fusion module and the corresponding sample label meet the constraints of the preset loss function.

[0079] Among them, the preset loss function can be used to determine multiple objectives for evaluating the prediction results, such as minimizing the prediction deviation, maximizing the compliance rate, etc. Reasonable weights can be set and the loss value can be calculated according to the specific demand scenario.

[0080] In this embodiment, the preset loss function includes two parts: prediction deviation loss and prediction result compliance rate loss. It can be expressed as: cost = w 1 *cost(error)+w 2 cost(qualified).

[0081] Among them, cost(error) represents the prediction deviation loss, which can be calculated and determined by the mean square error calculation method. The purpose of optimization through this loss calculation is to reduce the prediction deviation and fit the real curve data as much as possible:

[0082]

[0083] Cost (qualified) represents the loss of the prediction qualification rate. The optimization purpose is to improve the qualification rate and the volatility risk tolerance of the model prediction value:

[0084] cost(qualified)=1-num(qualified_days) / num(all_days)).

[0085] Among them, w 1 and w 2 is the weight parameter. The predicted value of the i-th day in an incoming line analysis cycle, is the true value of the sample label corresponding to the predicted value. num(qualified_days) indicates the number of days when the prediction result meets the standard, and num(all_days) indicates the total number of days in an incoming line analysis cycle.

[0086] The overall optimization goal of the preset loss function is to minimize the loss, which is expressed as: Minimize(w 1 *cost(error)+w2 cost(qualified)).

[0087] In addition, it should be noted that the model training samples are multi-dimensional feature data of each second-line customer service work group, that is, multiple feature data to be analyzed. The multiple feature data to be analyzed include multiple preset general feature data associated with historical incoming call volume data, and multiple preset work skill feature data associated with the professional skills of each work group. This setting is due to the fact that the incoming call volume distribution between different second-line customer service work groups is very different, the incoming calls are affected by many factors, the transmission links are long, and the correlation is weak.

[0088] The multiple preset common feature data associated with the historical incoming call volume data are common features of all second-line customer service workgroups, usually associated with the overall business volume, and have nothing to do with the business direction of a specific second-line customer service workgroup. The multiple preset work skill feature data associated with the professional skills of each workgroup are unique features of the corresponding second-line customer service workgroup, are strongly related to the business direction undertaken by the workgroup, and have a direct impact on the incoming call volume of the workgroup. In addition, when determining the work skill feature data, the business-specific features can be further subdivided according to the incoming line transmission link for the business undertaken by each second-line customer service workgroup, task category (category), and reception population.

[0089] Exemplarily, the preset common feature data may be feature data such as historical incoming call volume, date features, product / service related events, and front-line customer service data.

[0090] Among them, the historical incoming call volume can use the incoming call volume on the same day of the past few weeks (for example, Wednesdays of the past three weeks) as a feature to capture the seasonality and weekly periodicity of the incoming call volume. Date features can be the day of the year, the day of the month, the day of the week, whether it is a holiday, etc. These features can help the model capture the impact of a specific date or event on the incoming call volume. Product / service related events can be new product launches, large-scale promotional activities, service interruptions, etc. These events that are clearly defined in advance usually affect customers' demand for customer service. The data of front-line customer service can be information such as the resolution rate of front-line customer service and the ratio of forwarding to second-line customer service.

[0091] The preset work skill characteristic data may be data such as logistics time efficiency, labor transfer rate, dispute initiation rate, delivery time efficiency, proportion of work group incoming lines, and work order creation rate.

[0092] The above model training process is a multi-objective optimization process based on multi-feature data that is closer to the business chain logic of second-line customer service.

[0093] S320. Obtain multiple preset general feature data and multiple preset work skill feature data corresponding to each second-line customer service work group in the current incoming call volume analysis cycle.

[0094] S330. Input the plurality of preset general feature data and the plurality of preset work skill feature data into the target joint incoming call volume analysis and prediction model to obtain a plurality of groups of second-line customer service incoming call volume prediction results for each of the second-line customer service working groups.

[0095] S340. Based on the actual incoming call volume data of each second-line customer service working group in at least one adjacent historical incoming call volume analysis cycle, the corresponding multiple groups of second-line customer service incoming call volume prediction results are verified and analyzed to determine the corresponding target second-line customer service incoming call volume prediction results.

[0096] In a specific example, the target joint call volume analysis and prediction model trained by S310 was used to predict the call volume of second-line customer service. It was verified that the prediction deviation of the second-line customer service working group was reduced from 6% to 2%, and the number of customer service calls per day increased by 3.36, saving an average of 22.99 people per day. In terms of user experience dimension data, the proportion of 24-hour response increased by 10%, and the proportion of 72-hour completion increased by 8%.

[0097] The technical solution of the disclosed embodiment is to obtain a target joint incoming call volume analysis prediction model by training the model based on multi-dimensional feature data and a preset loss function including prediction deviation loss and prediction result compliance rate loss, and obtain a plurality of feature data to be analyzed for each second-line customer service working group in the current incoming call volume analysis cycle; input the plurality of feature data to be analyzed into the pre-trained joint incoming call volume analysis prediction model to obtain a plurality of second-line customer service incoming call volume prediction results for each second-line customer service working group; and verify and analyze the corresponding plurality of second-line customer service incoming call volume prediction results based on the actual incoming call volume data of each second-line customer service working group in at least one adjacent historical incoming call volume analysis cycle, and determine the corresponding target second-line customer service incoming call volume prediction results. The technical solution of the disclosed embodiment solves the problem of large deviation in the current prediction of the incoming call volume of second-line customer service with a longer business link length, and can predict the incoming call volume of second-line customer service by integrating multi-dimensional feature data affecting the incoming call volume, thereby reducing the prediction deviation, and further combining the historical incoming call volume data to determine the optimal prediction result among the plurality of prediction results.

[0098] Figure 4 An object incoming call volume analysis device is provided in an embodiment of the present disclosure. The device is suitable for scenarios where predictive analysis is performed on customer service incoming call volume, especially for accurately predicting the incoming call volume of second-line customer service with longer business links. The object incoming call volume analysis device can be implemented in the form of software and / or hardware and can be configured on an electronic device, which can be a mobile terminal, a PC or a server, etc.

[0099] like Figure 4 As shown, the object incoming line quantity analysis device includes: a feature data acquisition unit 410, an incoming line quantity prediction unit 420 and an incoming line quantity prediction result determination unit 430.

[0100] Among them, the feature data acquisition unit 410 is used to obtain multiple feature data to be analyzed for each target object set in the current incoming line quantity analysis cycle, and the target object set includes at least one first object, and the incoming line quantity of the first object has an associated relationship with at least one second object; the incoming line quantity prediction unit 420 is used to input the multiple feature data to be analyzed into a pre-trained joint incoming line quantity analysis prediction model to obtain multiple groups of first object incoming line quantity prediction results for each target object set; the incoming line quantity prediction result determination unit 430 is used to verify and analyze the corresponding multiple groups of first object incoming line quantity prediction results based on the actual incoming line quantity data of each target object set in at least one adjacent historical incoming line quantity analysis cycle, and determine the corresponding target first object incoming line quantity prediction result.

[0101] The technical solution of the disclosed embodiment is to obtain multiple feature data to be analyzed for each target object set in the current incoming line analysis cycle, wherein the target object set includes at least one first object, and the incoming line volume of the first object has an associated relationship with at least one second object; the multiple feature data to be analyzed are input into a pre-trained joint incoming line analysis prediction model to obtain multiple groups of first object incoming line prediction results for each target object set; based on the actual incoming line data of each target object set in at least one adjacent historical incoming line analysis cycle, the corresponding multiple groups of first object incoming line prediction results are verified and analyzed to determine the corresponding target first object incoming line prediction results. The technical solution of the disclosed embodiment solves the problem of large deviation in the current prediction of the incoming line volume for second-line customer service with longer service links, and can predict the incoming line volume of second-line customer service by integrating multi-dimensional feature data that affect the incoming line volume, thereby reducing the prediction deviation, and further combines the historical incoming line volume data to determine the optimal prediction result among multiple groups of prediction results.

[0102] In an optional implementation manner, the incoming line quantity prediction unit 420 is specifically used for:

[0103] Inputting the plurality of feature data to be analyzed into each independent incoming line quantity analysis and prediction module in the joint incoming line quantity analysis and prediction model respectively, to obtain a plurality of groups of first object incoming line quantity independent prediction results;

[0104] The multiple groups of independent prediction results of the first object incoming line quantity are input into the prediction result fusion module in the joint incoming line quantity analysis and prediction model to obtain the corresponding first object incoming line quantity fusion prediction results.

[0105] In an optional implementation, the incoming line quantity prediction unit 420 may also be used for:

[0106] Data enhancement processing of the incoming line quantity prediction result is performed according to the multiple groups of independent prediction results of the first object incoming line quantity and / or the fusion prediction results of the first object incoming line quantity to obtain multiple groups of amplified prediction results of the first object incoming line quantity.

[0107] In an optional implementation manner, the incoming line quantity prediction result determination unit 430 is specifically used to:

[0108] Respectively calculating the predicted deviation value between each group of the first object incoming line quantity prediction result and the incoming line quantity data corresponding to each time node in the actual incoming line quantity data;

[0109] Determining the compliance rate of each group of the first object incoming line quantity prediction results based on the prediction deviation value and the preset prediction compliance deviation standard;

[0110] A set of the first object incoming line quantity prediction results corresponding to the maximum compliance rate value is used as the corresponding target first object incoming line quantity prediction results.

[0111] In an optional implementation, the incoming line quantity prediction result determination unit 430 may also be used to:

[0112] Determining the number of prediction deviation values ​​that meet the target for the prediction result according to the prediction target deviation standard;

[0113] The ratio of the number of achieved targets to the total number of predicted deviation values ​​within an incoming line quantity analysis cycle is determined as the target achievement rate of each group of the first object incoming line quantity prediction results.

[0114] In an optional embodiment, the multiple feature data to be analyzed include: the multiple feature data to be analyzed include: multiple preset general feature data associated with historical incoming line quantity data and multiple preset work skill feature data associated with professional skills of each target object set.

[0115] In an optional implementation, the joint incoming line quantity analysis and prediction model includes a plurality of independent incoming line quantity analysis and prediction network modules and a prediction result fusion module;

[0116] Among them, the independent incoming line analysis and prediction network module includes at least two model structures of an autoregressive exponentially weighted average moving model, a time series model, an extreme gradient boosting multi-factor model, and a deep learning model with a preset structure.

[0117] In an optional implementation, the object incoming line quantity analysis device further includes a model training module, which is used to:

[0118] Before predicting the incoming line quantity of the target object set, model training is performed based on a preset loss function to obtain the joint incoming line quantity analysis and prediction model;

[0119] Among them, the preset loss function includes prediction deviation loss and prediction result compliance rate loss.

[0120] The object incoming line quantity analysis device provided in the embodiments of the present disclosure can execute the object incoming line quantity analysis method provided in any embodiment of the present disclosure, and has the corresponding functional modules and beneficial effects of the execution method.

[0121] It is worth noting that the various units and modules included in the above-mentioned device are only divided according to functional logic, but are not limited to the above-mentioned division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the protection scope of the embodiments of the present disclosure.

[0122] Figure 5 This is a schematic diagram of the structure of an electronic device provided by an embodiment of the present disclosure. Figure 5 , which shows an electronic device (eg, Figure 5 The terminal device in the embodiment of the present disclosure may include but is not limited to mobile terminals such as mobile phones, notebook computers, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 5 The electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0123] like Figure 5 As shown, the electronic device 500 may include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 501, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 502 or a program loaded from a storage device 508 into a random access memory (RAM) 503. In the RAM 503, various programs and data required for the operation of the electronic device 500 are also stored. The processing device 501, the ROM 502, and the RAM 503 are connected to each other via a bus 504. An edit / output (I / O) interface 505 is also connected to the bus 504.

[0124] Typically, the following devices may be connected to the I / O interface 505: an input device 506 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 507 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 508 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 509. The communication device 509 may allow the electronic device 500 to communicate with other devices wirelessly or by wire to exchange data. Although Figure 5 The electronic device 500 is shown with various devices, but it should be understood that it is not required to implement or possess all the devices shown. More or fewer devices may be implemented or possessed instead.

[0125] In particular, according to an embodiment of the present disclosure, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a computer program product, which includes a computer program carried on a non-transitory computer-readable medium, and the computer program contains program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through a communication device 509, or installed from a storage device 508, or installed from a ROM 502. When the computer program is executed by the processing device 501, the above-mentioned functions defined in the method of the embodiment of the present disclosure are executed.

[0126] The names of the messages or information exchanged between multiple devices in the embodiments of the present disclosure are only used for illustrative purposes and are not used to limit the scope of these messages or information.

[0127] The electronic device provided in the embodiment of the present disclosure and the object incoming line quantity analysis method provided in the above embodiment belong to the same inventive concept. The technical details not fully described in this embodiment can be referred to the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.

[0128] The embodiment of the present disclosure further provides a computer storage medium on which a computer program is stored. When the program is executed by a processor, the object incoming line quantity analysis method provided by the above embodiment is implemented.

[0129] It should be noted that the computer-readable medium disclosed above may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, device or device. In the present disclosure, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, in which a computer-readable program code is carried. This propagated data signal may take a variety of forms, including but not limited to an electromagnetic signal, an optical signal, or any suitable combination of the above. The computer readable signal medium may also be any computer readable medium other than a computer readable storage medium, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer readable medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0130] In some embodiments, the client and the server may communicate using any currently known or future developed network protocol such as HTTP (HyperText Transfer Protocol), and may be interconnected with any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network ("LAN"), a wide area network ("WAN"), an internet (e.g., the Internet), and a peer-to-peer network (e.g., an ad hoc peer-to-peer network), as well as any currently known or future developed network.

[0131] The computer-readable medium may be included in the electronic device, or may exist independently without being installed in the electronic device.

[0132] The computer-readable medium carries one or more programs. When the one or more programs are executed by the electronic device, the electronic device:

[0133] Acquire multiple pieces of feature data to be analyzed for each target object set in a current incoming line quantity analysis cycle, wherein the target object set includes at least one first object, and the incoming line quantity of the first object is associated with at least one second object;

[0134] Inputting the plurality of feature data to be analyzed into a pre-trained joint incoming line quantity analysis and prediction model to obtain a plurality of groups of first object incoming line quantity prediction results for each target object set;

[0135] Based on the actual incoming line quantity data of each target object set in at least one adjacent historical incoming line quantity analysis cycle, the corresponding multiple groups of first object incoming line quantity prediction results are verified and analyzed to determine the corresponding target first object incoming line quantity prediction results.

[0136] Computer program code for performing the operations of the present disclosure may be written in one or more programming languages ​​or a combination thereof, including, but not limited to, object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., through the Internet using an Internet service provider).

[0137] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present disclosure. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some implementations as replacements, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0138] The units involved in the embodiments described in the present disclosure may be implemented by software or hardware. The name of a unit does not limit the unit itself in some cases. For example, the first acquisition unit may also be described as a "unit for acquiring at least two Internet Protocol addresses".

[0139] The functions described above herein may be performed at least in part by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), and the like.

[0140] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0141] The embodiments of the present disclosure also provide a computer program product, including a computer program, which, when executed by a processor, implements the object incoming line quantity analysis method provided in any embodiment of the present disclosure.

[0142] In the process of implementation, the computer program product can be written in one or more programming languages ​​or a combination thereof to perform the computer program code for the disclosed operation, including object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computer (e.g., using an Internet service provider to connect through the Internet).

[0143] According to one or more embodiments of the present disclosure, [Example 1] provides a method for analyzing incoming line quantity of an object, the method comprising:

[0144] Acquire multiple pieces of feature data to be analyzed for each target object set in a current incoming line quantity analysis cycle, wherein the target object set includes at least one first object, and the incoming line quantity of the first object is associated with at least one second object;

[0145] Inputting the plurality of feature data to be analyzed into a pre-trained joint incoming line quantity analysis and prediction model to obtain a plurality of groups of first object incoming line quantity prediction results for each target object set;

[0146] Based on the actual incoming line quantity data of each target object set in at least one adjacent historical incoming line quantity analysis cycle, the corresponding multiple groups of first object incoming line quantity prediction results are verified and analyzed to determine the corresponding target first object incoming line quantity prediction results.

[0147] According to one or more embodiments of the present disclosure, [Example 2] provides a method for analyzing incoming line quantity of an object, further comprising:

[0148] In some optional implementations, the plurality of feature data to be analyzed are input into a pre-trained joint incoming line quantity analysis and prediction model to obtain a plurality of groups of first object incoming line quantity prediction results for each target object set, including:

[0149] Inputting the plurality of feature data to be analyzed into each independent incoming line quantity analysis and prediction module in the joint incoming line quantity analysis and prediction model respectively, to obtain a plurality of groups of first object incoming line quantity independent prediction results;

[0150] The multiple groups of independent prediction results of the first object incoming line quantity are input into the prediction result fusion module in the joint incoming line quantity analysis and prediction model to obtain the corresponding first object incoming line quantity fusion prediction results.

[0151] According to one or more embodiments of the present disclosure, [Example 3] provides a method for analyzing incoming line quantity of an object, including:

[0152] In some optional implementations, the process of obtaining multiple groups of first object incoming line quantity prediction results for each target object set further includes:

[0153] Data enhancement processing of the incoming line quantity prediction result is performed according to the multiple groups of independent prediction results of the first object incoming line quantity and / or the fusion prediction results of the first object incoming line quantity to obtain multiple groups of amplified prediction results of the first object incoming line quantity.

[0154] According to one or more embodiments of the present disclosure, [Example 4] provides a method for analyzing incoming line quantity of an object, further comprising:

[0155] In some optional implementations, based on actual incoming line quantity data of each target object set in at least one adjacent historical incoming line quantity analysis period, a verification analysis is performed on corresponding multiple groups of first object incoming line quantity prediction results to determine the corresponding target first object incoming line quantity prediction results, including:

[0156] Respectively calculating the predicted deviation value between each group of the first object incoming line quantity prediction result and the incoming line quantity data corresponding to each time node in the actual incoming line quantity data;

[0157] Determining the compliance rate of each group of the first object incoming line quantity prediction results based on the prediction deviation value and the preset prediction compliance deviation standard;

[0158] A set of the first object incoming line quantity prediction results corresponding to the maximum compliance rate value is used as the corresponding target first object incoming line quantity prediction results.

[0159] According to one or more embodiments of the present disclosure, [Example 5] provides a method for analyzing incoming line quantity of an object, further comprising:

[0160] In some optional implementations, determining the compliance rate of each group of the first object incoming line quantity prediction results based on the prediction deviation value and a preset prediction compliance deviation standard includes:

[0161] Determining the number of prediction deviation values ​​that meet the target for the prediction result according to the prediction target deviation standard;

[0162] The ratio of the number of achieved targets to the total number of predicted deviation values ​​within an incoming line quantity analysis cycle is determined as the target achievement rate of each group of the first object incoming line quantity prediction results.

[0163] According to one or more embodiments of the present disclosure, [Example 6] provides a method for analyzing incoming line quantity of an object, further comprising:

[0164] In some optional implementations, the plurality of feature data to be analyzed include: a plurality of preset general feature data associated with historical incoming line volume data and a plurality of preset work skill feature data associated with professional skills of each target object set.

[0165] According to one or more embodiments of the present disclosure, [Example 7] provides a method for analyzing incoming line quantity of an object, further comprising:

[0166] In some optional implementations, the joint incoming line quantity analysis and prediction model includes a plurality of independent incoming line quantity analysis and prediction network modules and a prediction result fusion module;

[0167] Among them, the independent incoming line analysis and prediction network module includes at least two model structures of an autoregressive exponentially weighted average moving model, a time series model, an extreme gradient boosting multi-factor model, and a deep learning model with a preset structure.

[0168] According to one or more embodiments of the present disclosure, [Example 8] provides a method for analyzing incoming line quantity of an object, further comprising:

[0169] In some optional implementations, the method further includes:

[0170] Before predicting the incoming line quantity of the target object set, model training is performed based on a preset loss function to obtain the joint incoming line quantity analysis and prediction model;

[0171] Among them, the preset loss function includes prediction deviation loss and prediction result compliance rate loss.

[0172] According to one or more embodiments of the present disclosure, [Example 9] provides an object incoming line quantity analysis device, including:

[0173] A feature data acquisition unit, used to acquire multiple feature data to be analyzed for each target object set in a current incoming line quantity analysis cycle, wherein the target object set includes at least one first object, and the incoming line quantity of the first object is associated with at least one second object;

[0174] An incoming line quantity prediction unit, used for inputting the plurality of feature data to be analyzed into a pre-trained joint incoming line quantity analysis prediction model to obtain a plurality of groups of first object incoming line quantity prediction results for each target object set;

[0175] The incoming line quantity prediction result determination unit is used to verify and analyze the corresponding multiple groups of the first object incoming line quantity prediction results based on the actual incoming line quantity data of each target object set in at least one adjacent historical incoming line quantity analysis cycle, and determine the corresponding target first object incoming line quantity prediction result.

[0176] According to one or more embodiments of the present disclosure, [Example 10] provides an object incoming line quantity analysis device, further comprising:

[0177] In an optional implementation manner, the incoming line quantity prediction unit is specifically used to:

[0178] Inputting the plurality of feature data to be analyzed into each independent incoming line quantity analysis and prediction module in the joint incoming line quantity analysis and prediction model respectively, to obtain a plurality of groups of first object incoming line quantity independent prediction results;

[0179] The multiple groups of independent prediction results of the first object incoming line quantity are input into the prediction result fusion module in the joint incoming line quantity analysis and prediction model to obtain the corresponding first object incoming line quantity fusion prediction results.

[0180] According to one or more embodiments of the present disclosure, [Example 11] provides an object incoming line quantity analysis device, further comprising:

[0181] In an optional implementation, the incoming line quantity prediction unit may also be used for:

[0182] Data enhancement processing of the incoming line quantity prediction result is performed according to the multiple groups of independent prediction results of the first object incoming line quantity and / or the fusion prediction results of the first object incoming line quantity to obtain multiple groups of amplified prediction results of the first object incoming line quantity.

[0183] According to one or more embodiments of the present disclosure, [Example 12] provides an object incoming line quantity analysis device, further comprising:

[0184] In an optional implementation manner, the incoming line quantity prediction result determination unit is specifically used to:

[0185] Respectively calculating the predicted deviation value between each group of the first object incoming line quantity prediction result and the incoming line quantity data corresponding to each time node in the actual incoming line quantity data;

[0186] Determining the compliance rate of each group of the first object incoming line quantity prediction results based on the prediction deviation value and the preset prediction compliance deviation standard;

[0187] A set of the first object incoming line quantity prediction results corresponding to the maximum compliance rate value is used as the corresponding target first object incoming line quantity prediction results.

[0188] According to one or more embodiments of the present disclosure, [Example 13] provides an object incoming line quantity analysis device, further comprising:

[0189] In an optional implementation, the incoming line quantity prediction result determination unit may also be used to:

[0190] Determining the number of prediction deviation values ​​that meet the target for the prediction result according to the prediction target deviation standard;

[0191] The ratio of the number of achieved targets to the total number of predicted deviation values ​​within an incoming line quantity analysis cycle is determined as the target achievement rate of each group of the first object incoming line quantity prediction results.

[0192] According to one or more embodiments of the present disclosure, [Example 14] provides an object incoming line quantity analysis device, further comprising:

[0193] In an optional embodiment, the multiple feature data to be analyzed include: the multiple feature data to be analyzed include: multiple preset general feature data associated with historical incoming line quantity data and multiple preset work skill feature data associated with professional skills of each target object set.

[0194] According to one or more embodiments of the present disclosure, [Example 15] provides an object incoming line quantity analysis device, further comprising:

[0195] In an optional implementation, the joint incoming line quantity analysis and prediction model includes a plurality of independent incoming line quantity analysis and prediction network modules and a prediction result fusion module;

[0196] Among them, the independent incoming line analysis and prediction network module includes at least two model structures of an autoregressive exponentially weighted average moving model, a time series model, an extreme gradient boosting multi-factor model, and a deep learning model with a preset structure.

[0197] According to one or more embodiments of the present disclosure, [Example 16] provides an object incoming line quantity analysis device, further comprising:

[0198] In an optional implementation, the object incoming line quantity analysis device further includes a model training module, which is used to:

[0199] Before predicting the incoming line quantity of the target object set, model training is performed based on a preset loss function to obtain the joint incoming line quantity analysis and prediction model;

[0200] Among them, the preset loss function includes prediction deviation loss and prediction result compliance rate loss.

[0201] The above description is only a preferred embodiment of the present disclosure and an explanation of the technical principles used. Those skilled in the art should understand that the scope of disclosure involved in the present disclosure is not limited to the technical solutions formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above disclosed concept. For example, the above features are replaced with the technical features with similar functions disclosed in the present disclosure (but not limited to) by each other to form a technical solution.

[0202] In addition, although each operation is described in a specific order, this should not be understood as requiring these operations to be performed in the specific order shown or in a sequential order. Under certain circumstances, multitasking and parallel processing may be advantageous. Similarly, although some specific implementation details are included in the above discussion, these should not be interpreted as limiting the scope of the present disclosure. Some features described in the context of a separate embodiment can also be implemented in a single embodiment in combination. On the contrary, the various features described in the context of a single embodiment can also be implemented in multiple embodiments individually or in any suitable sub-combination mode.

[0203] Although the subject matter has been described in language specific to structural features and / or methodological logical actions, it should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or actions described above. On the contrary, the specific features and actions described above are merely example forms of implementing the claims.

Claims

1. A method for analyzing incoming line quantity of an object, It is characterized in that include: Acquire multiple pieces of feature data to be analyzed for each target object set in a current incoming line quantity analysis cycle, wherein the target object set includes at least one first object, and the incoming line quantity of the first object is associated with at least one second object; Inputting the plurality of feature data to be analyzed into a pre-trained joint incoming line quantity analysis and prediction model to obtain a plurality of groups of first object incoming line quantity prediction results for each target object set; Based on the actual incoming line quantity data of each target object set in at least one adjacent historical incoming line quantity analysis cycle, the corresponding multiple groups of first object incoming line quantity prediction results are verified and analyzed to determine the corresponding target first object incoming line quantity prediction results.

2. The method according to claim 1, It is characterized in that Inputting the plurality of feature data to be analyzed into a pre-trained joint incoming line quantity analysis and prediction model to obtain a plurality of first object incoming line quantity prediction results for each target object set, including: Inputting the plurality of feature data to be analyzed into each independent incoming line quantity analysis and prediction module in the joint incoming line quantity analysis and prediction model respectively, to obtain a plurality of groups of first object incoming line quantity independent prediction results; The multiple groups of independent prediction results of the first object incoming line quantity are input into the prediction result fusion module in the joint incoming line quantity analysis and prediction model to obtain the corresponding first object incoming line quantity fusion prediction results.

3. The method according to claim 2, It is characterized in that The process of obtaining a plurality of first object incoming line quantity prediction results for each target object set also includes: Data enhancement processing of the incoming line quantity prediction result is performed according to the multiple groups of independent prediction results of the first object incoming line quantity and / or the fusion prediction results of the first object incoming line quantity to obtain multiple groups of amplified prediction results of the first object incoming line quantity.

4. The method according to claim 1, It is characterized in that Based on the actual incoming line quantity data of each target object set in at least one adjacent historical incoming line quantity analysis period, a verification analysis is performed on the corresponding multiple groups of the first object incoming line quantity prediction results to determine the corresponding target first object incoming line quantity prediction results, including: Respectively calculating the predicted deviation value between each group of the first object incoming line quantity prediction result and the incoming line quantity data corresponding to each time node in the actual incoming line quantity data; Determining the compliance rate of each group of the first object incoming line quantity prediction results based on the prediction deviation value and the preset prediction compliance deviation standard; A set of the first object incoming line quantity prediction results corresponding to the maximum compliance rate value is used as the corresponding target first object incoming line quantity prediction results.

5. The method according to claim 4, It is characterized in that Determining the compliance rate of each group of the first object incoming line quantity prediction results based on the prediction deviation value and the preset prediction compliance deviation standard includes: Determining the number of prediction deviation values ​​that meet the target for the prediction result according to the prediction target deviation standard; The ratio of the number of achieved targets to the total number of predicted deviation values ​​within an incoming line quantity analysis cycle is determined as the target achievement rate of each group of the first object incoming line quantity prediction results.

6. The method according to claim 1, It is characterized in that The plurality of feature data to be analyzed include: a plurality of preset general feature data associated with historical incoming line quantity data and a plurality of preset work skill feature data associated with professional skills of each target object set.

7. The method according to any one of claims 1 to 6, It is characterized in that The joint incoming line quantity analysis and prediction model includes a plurality of independent incoming line quantity analysis and prediction network modules and a prediction result fusion module; Among them, the independent incoming line analysis and prediction network module includes at least two model structures of an autoregressive exponentially weighted average moving model, a time series model, an extreme gradient boosting multi-factor model, and a deep learning model with a preset structure.

8. The method according to claim 7, It is characterized in that The method further comprises: Before predicting the incoming line quantity of the target object set, model training is performed based on a preset loss function to obtain the joint incoming line quantity analysis and prediction model; Among them, the preset loss function includes prediction deviation loss and prediction result compliance rate loss.

9. A device for analyzing incoming line quantity of an object, It is characterized in that include: A feature data acquisition unit, used to acquire multiple feature data to be analyzed for each target object set in a current incoming line quantity analysis cycle, wherein the target object set includes at least one first object, and the incoming line quantity of the first object is associated with at least one second object; An incoming line quantity prediction unit, used for inputting the plurality of feature data to be analyzed into a pre-trained joint incoming line quantity analysis prediction model to obtain a plurality of groups of first object incoming line quantity prediction results for each target object set; The incoming line quantity prediction result determination unit is used to verify and analyze the corresponding multiple groups of the first object incoming line quantity prediction results based on the actual incoming line quantity data of each target object set in at least one adjacent historical incoming line quantity analysis cycle, and determine the corresponding target first object incoming line quantity prediction result.

10. An electronic device, It is characterized in that The electronic device comprises: one or more processors; a storage device for storing one or more programs, When the one or more programs are executed by the one or more processors, the one or more processors implement the object incoming line quantity analysis method as described in any one of claims 1-8.

11. A computer-readable storage medium having a computer program stored thereon, It is characterized in that When the program is executed by a processor, the object incoming line quantity analysis method as described in any one of claims 1-8 is implemented.

12. A computer program product comprising a computer program, It is characterized in that When the computer program is executed by a processor, the computer program implements the object incoming line quantity analysis method according to any one of claims 1 to 8.

Images