An intelligent group call method, system, device and medium

Abstract

The application provides a kind of intelligent group call method, system, equipment and medium, it is related to outbound call technical field, and initial strategy parameter and right instruction are determined by responding call task through strategy decision module;Task scheduling module is based on call task, obtains call number through number resource module, generates call instruction in combination with the callable state feedback of communication execution module;Communication execution module executes call and returns result;Intelligent analysis module determines optimization strategy according to received data and feeds back to strategy decision module, to form "strategy configuration-execution monitoring-effect analysis-strategy optimization" closed loop, upgrade experience-driven strategy to data-driven strategy, realize the optimization efficiency of strategy while improving service efficiency.

Description

Technical Field

[0001] This invention relates to the field of outbound calling technology, and more specifically, to an intelligent group calling method, system, device, and medium. Background Technology

[0002] In outbound telephone technology, fixed-rate dialing or scheduling algorithms based on preset models are commonly used for concurrent call management. This method has the following drawbacks. For example, the scheduling strategy based on a fixed mathematical model calculates the optimal concurrency through preset parameters. However, this model relies on idealized assumptions about the communication scenario and has poor adaptability to dynamic disturbances in the actual environment. For instance, the quality of the number pool (such as the proportion of valid numbers and users' willingness to answer) can fluctuate drastically due to factors such as overdue periods and user attributes, resulting in a significant deviation between the actual connection rate and the model's predicted value. Existing fixed models cannot correct such deviations in real time, leading to scheduling errors that often exceed 25%, which in turn causes resource allocation imbalances, such as agent overload or idleness. Summary of the Invention

[0003] In view of this, the purpose of the present invention is to provide an intelligent group calling method, system, device and medium to improve service efficiency while optimizing strategy efficiency.

[0004] In a first aspect, this application provides an intelligent group calling system, including: a strategy decision-making module, a task scheduling module, a number resource module, a communication execution module, and an intelligent analysis module;

[0005] The strategy decision module is used to determine the initial call strategy parameters and number access control instructions corresponding to the call task in response to the call task operation.

[0006] The task scheduling module is used to determine the number allocation instruction corresponding to the call task based on the call task and initial call policy parameters; receive the call number corresponding to the call task; wherein the call number corresponding to the call task is obtained by the number resource module based on the number permission control instruction and the number allocation instruction; receive the callable status corresponding to the call number; wherein the callable status is obtained by the communication execution module based on the call number and the detection rules; and determine the call instruction corresponding to the call task based on the call number corresponding to the call task, the callable status of the call number, the call task, and the initial call policy parameters.

[0007] The communication execution module is used to receive the call instruction and call number corresponding to the call task, execute the call operation, receive the call result corresponding to the call operation and send it to the task scheduling module;

[0008] The intelligent analysis module is used to determine optimization strategies and send them to the strategy decision module based on the call tasks, initial call strategy parameters, callable status of the call number and call results received from the task scheduling module.

[0009] Optionally, the initial call strategy parameters include the regional concurrency coefficient, and the strategy decision module is also used for:

[0010] Obtain historical global call connection rate data and historical regional call connection rate;

[0011] Based on historical global connection rate data and historical regional connection rate, the regional concurrency coefficient is determined.

[0012] Optionally, the call instruction includes concurrent data; the task scheduling module is also used for:

[0013] Based on the call task, determine the corresponding concurrency base for the call task;

[0014] Based on the concurrency base and agent idle rate correction factor, determine the actual concurrency data corresponding to the call task;

[0015] Based on actual concurrent data and regional concurrent coefficients, determine the actual concurrent data for the target region.

[0016] Optionally, the task scheduling module is also used for:

[0017] Obtain the target connection rate data and actual connection rate data corresponding to the call task;

[0018] Based on the target connection rate data and the actual connection rate data, the actual concurrent data corresponding to the call task is optimized.

[0019] Optionally, the call instruction also includes an answering strategy; the task scheduling module is also used for:

[0020] Construct a bipartite graph and calculate the weight matrix of its edges; the bipartite graph is a bipartite graph of the call task and the agent; the weight matrix is ​​determined based on the answering strategy, agent status, and historical connection rate.

[0021] The optimal matching relationship between call tasks and agents is obtained by using the Hungarian algorithm based on the weight matrix.

[0022] Optionally, the task scheduling module is also used for:

[0023] When performing a call operation, a speech recognition model is used to extract and process the call content based on the call content to obtain a call summary corresponding to the call content.

[0024] Optionally, the number resource module is also used for:

[0025] Group and store call numbers, and / or;

[0026] Based on the connection rate of the number corresponding to the call number, the call number is assigned a corresponding tag.

[0027] Secondly, this application provides an intelligent group calling method, applied to the aforementioned intelligent group calling system, comprising:

[0028] In response to a call task operation, determine the initial call policy parameters and number access control instructions corresponding to the call task;

[0029] Based on the call task and initial call policy parameters, determine the number allocation instruction corresponding to the call task;

[0030] The calling number corresponding to the calling task is determined based on the number access control instruction and the number allocation instruction;

[0031] The call availability status of a call number is determined based on the call number and detection rules.

[0032] Based on the call number corresponding to the call task, the call availability status of the call number, the call task and the initial call strategy parameters, determine the call instruction corresponding to the call task;

[0033] Based on the call instruction and call number corresponding to the call task, execute the call operation and receive the call result corresponding to the call operation;

[0034] Based on the call task, initial call strategy parameters, callability status of the call number, and call results, an optimization strategy is determined.

[0035] Thirdly, this application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement an intelligent group calling method.

[0036] Fourthly, this application also provides a computer-readable storage medium storing computer instructions, which, when executed by a processor, implement the intelligent group calling method.

[0037] This application provides an intelligent group calling method, system, device, and medium. A strategy decision module responds to call tasks, determining initial strategy parameters and permission instructions. A task scheduling module, based on the call task, obtains the call number via a number resource module and, combined with the callability status feedback from the communication execution module, generates a call instruction. The communication execution module executes the call and transmits the result back. An intelligent analysis module determines an optimization strategy based on received data and feeds it back to the strategy decision module, forming a closed loop of "strategy configuration - execution monitoring - effect analysis - strategy optimization." This upgrades experience-driven strategies to data-driven strategies, improving service efficiency while optimizing strategy efficiency.

[0038] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0039] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0040] Figure 1 A flowchart of an intelligent group calling method provided by an embodiment of the present invention is shown;

[0041] Figure 2 A flowchart of an intelligent group calling system provided by an embodiment of the present invention is shown;

[0042] Figure 3 A schematic diagram of the structure of an electronic device provided in an embodiment of the present invention is shown. Detailed Implementation

[0043] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0044] This application provides an intelligent group calling system, see below. Figure 1 As shown, the intelligent group calling system provided in this application embodiment includes a strategy decision module 110, a task scheduling module 120, a number resource module 130, a communication execution module 140, and an intelligent analysis module 150;

[0045] The strategy decision module 110 is used to determine the initial call strategy parameters and number access control instructions corresponding to the call task in response to the call task operation.

[0046] In this embodiment, the call strategy parameters include, but are not limited to, concurrency control-related parameters and task scheduling adaptation parameters. The concurrency control-related parameters include, but are not limited to, regional concurrency coefficient, concurrency base, actual concurrency, and concurrency adjustment. The task scheduling adaptation parameters include, but are not limited to, agent parameters and answering strategies. The answering strategy includes multi-priority collaborative answering (i.e., a four-level flexible answering strategy), response time threshold, and priority switching delay. For example, the four-level flexible answering strategy could have a first priority of dedicated mediator, a second priority of backup mediators within the same group, a third priority of cross-group mediators, and a fourth priority of a voice recognition robot. The response time threshold sets the maximum response waiting time for the first or current priority, such as 150ms. When the maximum response waiting time is exceeded, the next priority switch is triggered to control the response efficiency of a single call. The priority switching delay limits the maximum time for priority switching (≤300ms) to ensure the real-time performance of multi-priority collaboration and prevent call failure due to slow switching.

[0047] The multi-order collaborative answering method provided in this application embodiment solves the problem of wasted agent resources in traditional single-threaded outbound calls by scheduling through response time thresholds, balancing agent workload and response time, thereby improving overall outbound call efficiency.

[0048] In this embodiment of the application, the strategy decision module 110 is further used to obtain historical global connection rate data and historical regional connection rate; and to determine the regional concurrency coefficient based on the historical global connection rate data and historical regional connection rate.

[0049] In a specific embodiment of this application, the strategy decision module 110 is further used to adjust the regional concurrency coefficient in real time using a Bayesian algorithm and a sliding window. The specific process is as follows:

[0050] The number of successful calls and the total number of global call attempts are obtained by using a preset sliding window step size.

[0051] Based on the historical number of successfully connected calls and the total number of call attempts, the historical global connection rate data is determined. The mathematical expression for the historical global connection rate data is as follows:

[0052]

[0053] In the formula, This represents the historical global connection rate. The total number of successfully connected calls globally. This represents the total number of global call attempts.

[0054] Historical area call connection rate data is determined based on the number of successful calls connected in the area and the total number of call attempts in the area.

[0055] Based on historical global call connection rate data and historical regional call connection rate data, the regional concurrency coefficient is determined, and its mathematical expression is as follows:

[0056]

[0057] In the formula, This is the regional concurrency coefficient. For regional connection rate, Historical global connection rate;

[0058] The Bayesian algorithm is used to optimize the regional concurrency coefficient in real time to obtain the target regional concurrency coefficient.

[0059] In this embodiment of the application, the strategy decision module 110 is also used to determine the number access control instruction in response to the call task operation. The number access control can be determined by a role-based access control (RBAC) model, such as allowing only Department A to call numbers in the East China region.

[0060] The task scheduling module 120 is used to determine the number allocation instruction corresponding to the call task based on the call task and the initial call policy parameters; receive the call number corresponding to the call task; wherein the call number corresponding to the call task is obtained by the number resource module 130 based on the number permission control instruction and the number allocation instruction; receive the callable status corresponding to the call number; wherein the callable status is obtained by the communication execution module 140 based on the call number and the detection rules; and determine the call instruction corresponding to the call task based on the call number corresponding to the call task, the callable status of the call number, the call task, and the initial call policy parameters; wherein the callable status of the number is determined by the communication execution module 140.

[0061] The communication execution module 140 is used to receive the call instruction and call number corresponding to the call task, execute the call operation, receive the call result corresponding to the call operation and send it to the task scheduling module 120.

[0062] The intelligent analysis module 150 is used to determine an optimization strategy and send it to the strategy decision module 110 based on the call task, initial call strategy parameters, callable status of the call number and call result received from the task scheduling module 120.

[0063] In the intelligent group calling system provided in this application embodiment, the strategy decision module 110 dynamically acquires historical call data using a sliding window and optimizes the regional concurrency coefficient in real time using a Bayesian algorithm. This allows the regional concurrency control parameters to accurately adapt to the call connection rate characteristics and real-time changes of different regions, avoiding the problem of resource allocation imbalance under traditional fixed concurrency strategies. This effectively improves the utilization rate of global call resources, thereby enhancing the overall call task connection rate and execution efficiency. The strategy decision module 110 determines number permission control instructions based on the RBAC model, through "user-role-permission". The associated mechanism precisely restricts the scope of number calls for different departments and roles, preventing unauthorized use of number resources, ensuring the security of number information, meeting compliance requirements in different business scenarios, and reducing operational risks caused by permission confusion. When generating call instructions, the task scheduling module 120 comprehensively considers multi-dimensional information such as call task requirements, initial policy parameters, and callable status of numbers, ensuring that call instructions are executed only for numbers that meet the permission requirements and are in a callable state. Combined with the real-time detection of callable status by the communication execution module 140, invalid calls are significantly reduced, communication resource waste is reduced, and the execution quality of call tasks is improved. The intelligent analysis module 150 generates optimization strategies based on full data during the call task execution process and feeds them back to the policy decision module 110, realizing closed-loop management of "policy formulation - task execution - result analysis - policy optimization". This allows it to adapt to changes in call demand in different business scenarios and at different times, maintaining a highly efficient and stable operating state in the long term.

[0064] In this embodiment, the call instruction includes, but is not limited to, basic communication parameters, policy execution parameters, service quality control data, and service context information. The basic communication parameters include the calling number, the called number, and the call protocol. The policy execution parameters include concurrency data, concurrency control ID, risk level label, and priority identifier. The service quality control data includes agent ID and skill group, expected threshold of MOS value, and maximum ringing duration. The service context information includes task ID, associated work order number, and customer label.

[0065] In this embodiment of the application, the task scheduling module 120 is further configured to determine the concurrency base corresponding to the call task based on the call task; determine the actual concurrency data corresponding to the call task based on the concurrency base and the agent idle rate correction factor; and determine the actual concurrency data of the target area based on the actual concurrency data and the regional concurrency coefficient.

[0066] In this embodiment of the application, the strategy decision module 110 is further configured to: obtain target connection rate data and actual connection rate data corresponding to the call task; and optimize the actual concurrent data corresponding to the call task based on the target connection rate data and the actual connection rate data.

[0067] In a specific embodiment of this application, the mathematical expression for the concurrency cardinality is:

[0068]

[0069] In the formula, As the concurrency base, For connectivity rate, For dynamic coefficients, .

[0070] The connectability rate is the percentage of all numbers included in each case, such as phone numbers and contact information, that can be successfully contacted after a connectability test, relative to the total number of numbers tested in that case.

[0071] The actual concurrent data expression:

[0072]

[0073]

[0074] In the formula, For actual concurrent data, This is a correction factor for seat vacancy rate. Number of available seats Total number of seats;

[0075] The actual concurrent data corresponding to the call task is optimized using PID as follows:

[0076]

[0077]

[0078] In the formula, This is the actual number of concurrent users adjusted. This is the error value. To achieve the target call connection rate, This represents the actual connection rate. , and This is the adjustment coefficient.

[0079] The above-described triple control method of error feedback, integral adjustment, and derivative prediction provided in this application embodiment determines the actual number of concurrent connections, thereby achieving millisecond-level dynamic optimization of the number of concurrent connections.

[0080] In this embodiment of the application, the task scheduling module 120 is further configured to: construct a bipartite graph and calculate the weight matrix of the edges of the bipartite graph; wherein, the bipartite graph is a bipartite graph of the call task and the agent; the weight matrix is ​​determined based on the answering strategy, the agent status and the historical connection rate; and the optimal matching relationship between the call task and the agent is obtained by calculating the weight matrix using the Hungarian algorithm.

[0081] In a specific embodiment of this application, firstly, nodes of a bipartite graph of call tasks and agents are defined, wherein the left node set is all call tasks to be assigned, such as call task 1, call task 2, ... call task n, and each call task includes target customer type, required skill tags and / or call priority, etc.; the right node set is all available agents, such as agent A, agent B, ..., agent m, and each agent includes skill range, current working status and / or historical performance, etc.

[0082] Secondly, define the edges of the bipartite graph of call tasks and agents. In the bipartite graph, each condition connects a call task and an agent, that is, the agent has the basic ability to handle the task. If the agent is completely unable to handle the task, no edge is created.

[0083] Then, based on the established bipartite graph of call tasks and agents, a weight matrix is ​​calculated (i.e., the weights of the edges are assigned), where the weight matrix is ​​a... The matrix, For the number of tasks, Let W[i][j] represent the number of seats, and let W[i][j] represent the matching weight assigned to seat j for task i. A higher matching weight indicates a better match. The mathematical expression for the weight matrix is:

[0084]

[0085] In the formula, This is the weight matrix. Let i be the priority coefficient for calling task i. The historical connection rate of agent j. The current load rate of agent j; , and For coefficients;

[0086] Among them, the priority coefficient of the call task reflects the urgency or importance of the call task. For example, high priority tasks need to be assigned first, and the value is usually [0,1]. The historical connection rate of the agent is the proportion of successful connection when the agent has handled similar call tasks in the past, and the value is usually [0,1]. The current load rate of the agent is the proportion of the amount of tasks that the agent is currently handling to its maximum capacity, reflecting the busyness of the agent, and the value is usually [0,1]. , and The coefficients can be adjusted according to business scenarios, such as periods of high turnover or resource scarcity, to better match call tasks with agents and ensure that the actual needs are met. , and The constraints on the coefficients are: + + =1;

[0087] Finally, the Hungarian algorithm is used to find a matching in the bipartite graph that maximizes the total weight or minimizes the total cost of the matching, ensuring that each task matches only one agent and each agent matches at most one task. First, the weight matrix is ​​converted into a cost matrix. Then, the minimum value of each row in the cost matrix is ​​subtracted to ensure that each row contains at least one 0, indicating that the position in that row is the optimal choice. The minimum value of each column in the processed cost matrix is ​​subtracted to ensure that each column contains at least one 0. All 0 elements are covered with the fewest possible straight lines (e.g., horizontal or vertical). If the number of straight lines is... If the number of lines equals the number of tasks (or the number of seats, whichever is smaller), then the optimal match has been found; otherwise, find the minimum value among the elements not covered by a line, subtract this value from all uncovered elements, add this value to the elements covered by two lines, and repeat covering all 0 elements with the fewest lines (such as horizontal or vertical). If the number of lines equals the number of tasks (or the number of seats, whichever is smaller), then the optimal match has been found. Continue until the optimal solution is found, and obtain the optimal matching relationship corresponding to the position of the 0 element in the matrix. That is, the match between calling task i and seat j is the optimal solution. At this time, the total weight of all matches is maximized.

[0088] The embodiments of this application determine the optimal matching relationship between call tasks and agents through the above method, thereby optimizing the overall matching efficiency and improving the call center's connection rate and resource utilization.

[0089] In this application embodiment, the callable status includes normal callable, conditional callable, and uncallable. Uncallable includes basic uncallable, risk-based uncallable, and service-based uncallable. Conditional callable means that a call can only be made if preset conditions are met, such as secondary confirmation of callable status or time-limited callable status. Basic uncallable means that the number cannot be connected due to its own status, such as being suspended, invalid, or canceled. Risk-based uncallable means that the call cannot be made due to rule blocking, such as numbers with high complaint rates, blacklisted numbers, and frequently rejected numbers. Service-based uncallable means that the call is blocked due to mismatched business scenarios, such as incompatible regions or incompatible number types (the task requires a landline, but it is actually a mobile phone number).

[0090] In this embodiment, the callable status of a call number is determined by the communication execution module 140 based on the call number and detection rules. The validity of the call number can be determined by detecting whether the call number format is correct and verifying whether the call number is an empty number or has been cancelled. The validity of the call number can also be determined by checking whether the call number is in the enterprise's custom blacklist or in the industry's shared blacklist.

[0091] In this embodiment, the call result includes basic identification parameters, status result parameters, answering parameters, and abnormal parameters; the basic identification parameters include the call task ID, the call number, and the call time; the status result parameters include the call status and the connection indicator; the answering parameters include the order execution status and the order switching information; the abnormal parameters include the call failure reason, the call duration, and additional information, such as abnormal prompts during the call and user feedback markers.

[0092] In this embodiment, the communication execution module 140 constructs a real-time communication gateway based on WebRTC technology, which supports bidirectional conversion between SIP protocol and PSTN network to ensure that the call quality MOS value reaches 4.2+.

[0093] In this embodiment of the application, the task scheduling module 120 is further configured to: when performing a call operation, extract and process the call content based on the call content using a speech recognition model to obtain a call summary corresponding to the call content.

[0094] In a specific embodiment of this application, a speech recognition model is used to extract and process the speech signal of the call content to determine the call summary corresponding to the content. The speech recognition model includes an input layer, a long short-term memory network layer, an attention mechanism layer, and an output layer. The input layer receives the speech signal and converts it into a sequence of spectral feature vectors through framing, windowing, and Fourier transform. The long short-term memory network layer receives the feature vector sequence from the input layer and extracts the semantics of each frame of speech through the synergistic effect of the forget gate, input gate, and output gate. The attention mechanism layer uses attention weights to weightedly sum the value vectors to obtain a context vector that incorporates key information. The output layer uses a fully connected layer and an activation function to convert the context vector into a probability distribution for each speech category, thus obtaining the call summary corresponding to the call content.

[0095] The speech recognition model with bidirectional LSTM+Attention mechanism provided in this application embodiment has a word error rate of less than 8.7%, supports real-time generation of call summaries, and achieves a summary completeness of 92%.

[0096] In a specific embodiment of this application, the number resource module 130 is further configured to: group and store the calling numbers, and / or: assign a corresponding tag to the calling number based on the connection rate of the number corresponding to the calling number.

[0097] Specifically, the number resource module 130 stores numbers from different departments in independent data shards; and predicts the number connection rate based on a neural network model, automatically marking numbers with predicted values ​​less than a threshold (such as 0.3) as high-risk numbers, so as to achieve full lifecycle security management of number resources, which improves security by more than 80% compared with the existing single-layer management mode.

[0098] In a scenario where a bank's credit card center needs to collect overdue payments on 100,000 accounts, the intelligent group calling system provided in this application can complete the collection within 4.2 days, whereas the traditional system would take 15 days. The effective connection rate increases from 13.2% to 31.5%, and the recovered amount commission increases by 220%. The outbound call configuration scheme in the group calling system provided in this application uses the following dynamic formula for concurrent calls: N = Connectivity Rate × 100 × 0.55; the priority rule is: first priority is collection specialist, second priority is team leader, and third priority is AI robot; the redial strategy is 3 calls per day. The call schedule is 4 hours apart. During peak sales periods (such as Double 11), when customer inquiries surge, traditional systems have a missed call rate exceeding 30% in e-commerce customer service follow-up scenarios. However, the group calling system provided in this application controls the missed call rate to within 5.8%, increases customer satisfaction from 72% to 91%, and reduces agent manpower costs by 35%. The outbound call configuration scheme in the group calling system provided in this application is as follows: enabling dual-dimensional number pool segmentation based on region and time period, adopting dynamic optimization of priority order, and having AI robots handle 40% of simple inquiries.

[0099] This application provides an intelligent group calling method, see below. Figure 2 As shown, the intelligent group calling method provided in this application includes:

[0100] Step 210: In response to the call task operation, determine the initial call policy parameters and number access control instructions corresponding to the call task;

[0101] Step 220: Based on the call task and initial call policy parameters, determine the number allocation instruction corresponding to the call task;

[0102] Step 230: Determine the calling number corresponding to the calling task based on the number access control instruction and the number allocation instruction;

[0103] Step 240: Determine the callability status of the call number based on the call number and detection rules;

[0104] Step 250: Based on the call number corresponding to the call task, the call availability status of the call number, the call task, and the initial call strategy parameters, determine the call instruction corresponding to the call task;

[0105] Step 260: Based on the call instruction and call number corresponding to the call task, execute the call operation and receive the call result corresponding to the call operation;

[0106] Step 270: Determine the optimization strategy based on the call task, initial call strategy parameters, callable status of the call number, and call results.

[0107] It should be noted that the principle of the intelligent group calling method provided in this application embodiment to solve the technical problem is similar to that of the intelligent group calling system provided in this application embodiment. Therefore, the implementation of the intelligent group calling method provided in this application embodiment can refer to the implementation of the intelligent group calling system provided in this application embodiment, and the repeated parts will not be described again.

[0108] After introducing the intelligent group calling system and method provided in the embodiments of this application, the electronic device provided in the embodiments of this application will be briefly introduced next.

[0109] See Figure 3 As shown, the electronic device 500 provided in this application embodiment includes at least a processor 501, a memory 502, and a computer program stored in the memory 502 and executable on the processor 501. When the processor 501 executes the computer program, it implements the intelligent group calling method provided in this application embodiment.

[0110] The electronic device 500 provided in this application embodiment may further include a bus 503 connecting different components (including processor 501 and memory 502). The bus 503 represents one or more types of bus structures, including memory bus, peripheral bus, local area bus, etc.

[0111] Memory 502 may include a readable storage medium in the form of volatile memory, such as random access memory (RAM) 5021 and / or cache memory 5022, and may further include read-only memory (ROM) 5023. Memory 502 may also include a program tool 5025 having a set (at least one) of program modules 5024, including but not limited to an operating subsystem, one or more application programs, other program modules, and program data, each or some combination of these examples may include an implementation of a network environment.

[0112] Processor 501 can be a single processing element or a collective term for multiple processing elements. For example, processor 501 can be a central processing unit (CPU) or one or more integrated circuits configured to implement the intelligent group calling method provided in the embodiments of this application. Specifically, processor 501 can be a general-purpose processor, including but not limited to CPUs, application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.

[0113] Electronic device 500 can communicate with one or more external devices 504 (e.g., keyboard, remote control, etc.), and also with one or more devices that enable a user to interact with electronic device 500 (e.g., mobile phone, computer, etc.), and / or with devices that enable electronic device 500 to communicate with one or more other electronic devices 500 (e.g., router, modem, etc.). This communication can be performed through input / output (I / O) interface 505. Furthermore, electronic device 500 can also communicate with one or more networks (e.g., local area network (LAN), wide area network (WAN), and / or public networks, such as the Internet) through network adapter 506. Figure 3 As shown, network adapter 506 communicates with other modules of electronic device 500 via bus 503. It should be understood that, although... Figure 3 As not shown, other hardware and / or software modules may be used in conjunction with the electronic device 500, including but not limited to microcode, device drivers, redundant processors, external disk drive arrays, Redundant Arrays of Independent Disks (RAID) subsystems, tape drives, and data backup storage subsystems.

[0114] It should be noted that, Figure 3 The electronic device 500 shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this application.

[0115] The following describes the computer-readable storage medium provided in the embodiments of this application. The computer-readable storage medium provided in the embodiments of this application stores computer instructions, which, when executed by a processor, implement the intelligent group calling method provided in the embodiments of this application. Specifically, the computer instructions can be built into or installed in the processor, so that the processor can implement the intelligent group calling method provided in the embodiments of this application by executing the built-in or installed computer instructions.

[0116] In addition, the intelligent group calling method provided in the embodiments of this application can also be implemented as a computer program product, which includes program code. The program code implements the intelligent group calling method provided in the embodiments of this application when it runs on a processor.

[0117] The computer program product provided in this application embodiment may employ one or more computer-readable storage media, which may be, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination thereof. Specifically, more specific examples (a non-exhaustive list) of computer-readable storage media include electrical connections with one or more wires, portable disks, hard disks, RAM, ROM, erasable programmable read-only memory (EPROM), optical fibers, portable compact disc read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0118] The computer program product provided in this application embodiment can be a CD-ROM and include program code, and can also run on electronic devices such as computers. However, the computer program product provided in this application embodiment is not limited thereto. In this application embodiment, the computer-readable storage medium can be any tangible medium that contains or stores program code, which can be used by or in conjunction with an instruction execution system, device, or apparatus.

[0119] It should be noted that although several units or sub-units of the device have been mentioned in the detailed description above, this division is merely exemplary and not mandatory. In fact, according to embodiments of this application, the features and functions of two or more units described above can be embodied in one unit. Conversely, the features and functions of one unit described above can be further divided and embodied by multiple units.

[0120] Furthermore, although the operations of the method of this application are described in a specific order in the accompanying drawings, this does not require or imply that these operations must be performed in that specific order, or that all the operations shown must be performed to achieve the desired result. Additionally or alternatively, certain steps may be omitted, multiple steps may be combined into one step, and / or one step may be broken down into multiple steps.

[0121] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0122] Obviously, those skilled in the art can make various modifications and variations to the embodiments of this application without departing from the spirit and scope of the embodiments of this application. Therefore, if these modifications and variations to the embodiments of this application fall within the scope of the claims of this application and their equivalents, this application also intends to include these modifications and variations.

Claims

1. An intelligent group call system, characterized by Comprise: A policy decision module, a task scheduling module, a number resource module, a communication execution module and an intelligent analysis module; The policy decision module is configured to determine initial call policy parameters and number authority control instructions corresponding to the call task in response to the call task operation; the initial call policy parameters include a regional concurrency coefficient; the policy decision module is further configured to: acquire historical global connection rate data and historical regional connection rates; determine the regional concurrency coefficient based on the historical global connection rate data and the historical regional connection rates; and adjust the regional concurrency coefficient in real time through a Bayesian algorithm and a sliding window, including: acquiring the number of successful connection calls and the total number of call attempts in the global and each region through a preset sliding window step; determining the historical global connection rate data based on the number of successful connection calls in the global and the total number of call attempts in the global; determining the historical regional connection rate data based on the number of successful connection calls in the region and the total number of call attempts in the region; and determining the regional concurrency coefficient based on the historical global connection rate data and the historical regional connection rate data, with a mathematical expression of the regional concurrency coefficient being: , wherein, is the regional concurrency coefficient, is the regional connection rate, is the historical global connection rate; and the regional concurrency coefficient is optimized in real time through a Bayesian algorithm to obtain a target regional concurrency coefficient. The task scheduling module is used for determining a number allocation instruction corresponding to the call task based on the call task and the initial call strategy parameter; receiving a call number corresponding to the call task; wherein the call number corresponding to the call task is obtained by the number resource module based on the number permission control instruction and the number allocation instruction; receiving a callable state corresponding to the call number; wherein the callable state is obtained by the communication execution module based on the call number and a detection rule; determining a call instruction corresponding to the call task based on the call number corresponding to the call task, the number callable state corresponding to the call number, the call task and the initial call strategy parameter; The communication execution module is used for receiving the call instruction and the call number corresponding to the call task, and executing a call operation; receiving a call result corresponding to the call operation and sending it to the task scheduling module; The intelligent analysis module is used for determining an optimization strategy based on the call task, the initial call strategy parameter, the call number callable state and the call result received by the task scheduling module, and sending it to the policy decision module.

2. The intelligent group call system of claim 1, wherein, The call instruction comprises concurrent data; the task scheduling module is further used for: Determining an actual concurrent data corresponding to the call task based on the concurrent base and a seat idle rate correction factor; Determining an actual concurrent data corresponding to the call task based on the actual concurrent data and the regional concurrent coefficient. The task scheduling module is further used for:

3. The intelligent group call system of claim 2, wherein, Obtaining target connection rate data and actual connection rate data corresponding to the call task; Optimizing the actual concurrent data corresponding to the call task based on the target connection rate data and the actual connection rate data. The call instruction further comprises an answering strategy; the task scheduling module is further used for:

4. The intelligent group call system of claim 3, wherein, Building a bipartite graph and calculating a weight matrix of edges of the bipartite graph; wherein the bipartite graph is a bipartite graph of the call task and a seat; the weight matrix is determined based on the answering strategy, a seat state and a historical connection rate; Calculating the optimal matching relationship between the call task and the seat based on the weight matrix by using the Hungarian algorithm. The task scheduling module is further used for:

5. The intelligent group call system of claim 1, wherein, When executing the call operation, extracting and processing the call content by using a speech recognition model based on the call content, to obtain a call summary corresponding to the call content. The number resource module is further used for:

6. The intelligent group call system of claim 1, wherein, Grouping and storing the call number, and / or; Assigning a corresponding mark to the call number based on the number connection rate corresponding to the call number. Applied to the intelligent group call system in any one of claims 1 to 6, comprising:

7. A method of intelligent group call, characterized by, Determining a number allocation instruction corresponding to the call task based on the call task and the initial call strategy parameter; In response to the call task operation, an initial call strategy parameter and a number authority control instruction corresponding to the call task are determined; the initial call strategy parameter comprises a regional concurrency coefficient, and the strategy decision module is further configured to: acquire historical global connection rate data and historical regional connection rates; based on the historical global connection rate data and the historical regional connection rates, the regional concurrency coefficient is determined; wherein the regional concurrency coefficient is adjusted in real time through a Bayesian algorithm and a sliding window, comprising: acquiring, through a preset sliding window step, a number of successful connection calls and a total number of call attempts in a global and in each region; based on the number of successful connection calls in the global and the total number of call attempts in the global, the historical global connection rate data is determined; based on the number of successful connection calls in the region and the total number of call attempts in the region, the historical regional connection rate data is determined; based on the historical global connection rate data and the historical regional connection rate data, the regional concurrency coefficient is determined, and a mathematical expression of the regional concurrency coefficient is: , wherein, is the regional concurrency coefficient, is the regional connection rate, is the historical global connection rate; the Bayesian algorithm is used to optimize the regional concurrency coefficient in real time to obtain a target regional concurrency coefficient; ​ determine a call number corresponding to the call task based on the number permission control instruction and the number allocation instruction; determine a callable state of the call number based on the call number and a detection rule; determine a call instruction corresponding to the call task based on the call number corresponding to the call task, the callable state of the call number, the call task and the initial call strategy parameter; perform a call operation based on the call instruction corresponding to the call task and the call number, and receive a call result corresponding to the call operation; determine an optimization strategy based on the call task, the initial call strategy parameter, the callable state of the call number and the call result.

8. An electronic device, comprising: The computer readable storage medium stores computer instructions, and the computer instructions are executed by the processor to implement the intelligent group call method in claim 7.

9. A computer-readable storage medium, characterized in that, The computer readable storage medium stores computer instructions, and the computer instructions are executed by the processor to implement the intelligent group call method in claim 7.

Images