Intelligent business opportunity business processing optimization method and system

Abstract

The invention discloses an intelligent business opportunity business processing optimization method and system, and relates to the technical field of artificial intelligence, and the method comprises the steps: collecting historical business opportunity data, and carrying out the multi-dimensional feature extraction, and obtaining business opportunity feature data; constructing a time sequence prediction network based on an attention mechanism, and performing data processing on the business opportunity feature data to obtain business opportunity conversion probability data; generating a business opportunity allocation scheme through a reinforcement learning algorithm according to the business opportunity conversion probability data in combination with sales team capability data and resource utilization data; and monitoring the execution state of the business opportunity allocation scheme, and when it is detected that a state index deviates from a preset threshold value, outputting early warning information and an intervention suggestion. The business opportunity processing efficiency and the conversion rate are improved, and reliable technical support is provided for digital transformation of enterprise sales management.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to an intelligent business opportunity service processing optimization method and system. Background Art

[0002] With the deepening of digital transformation of enterprises, business opportunity management, as a core link of sales business, has a direct impact on the market competitiveness of enterprises through its intelligence level. Traditional business opportunity management systems mainly rely on manual experience to identify and allocate business opportunities, usually using simple rule matching or polling allocation methods. In recent years, with the development of machine learning technology, some companies have begun to try to introduce predictive analysis models to assist business opportunity management decisions, such as regression analysis and classification algorithms based on historical data. At the same time, the breakthrough of deep learning in time series data processing has significantly improved the prediction accuracy of business opportunity conversion probability. In particular, the introduction of the attention mechanism can effectively capture the key time series characteristics in the process of business opportunity development, providing new technical support for intelligent business opportunity management.

[0003] However, there are still several key problems with existing business opportunity management technologies: First, most systems only focus on single-dimensional feature analysis and fail to fully integrate multi-dimensional information such as customer portraits, market trends, and competitive situations, resulting in incomplete extraction of business opportunity features; second, existing business opportunity allocation plans often use static rule engines, lack the ability to dynamically adapt to sales team capabilities and resource utilization, and cannot achieve optimal resource allocation; third, the business opportunity execution process lacks an effective monitoring mechanism, making it difficult to detect and intervene in abnormal situations in a timely manner, affecting the effect of business opportunity conversion. In addition, existing technologies generally have problems such as poor model interpretability and opaque decision-making basis, which limits the application effect of intelligent business opportunity management systems in actual business scenarios.

[0004] In summary, the existing business opportunity processing technology has obvious deficiencies in feature engineering, resource allocation optimization, process monitoring, etc., and there is an urgent need for a technical solution that can realize intelligent business opportunity feature extraction, dynamic optimization allocation decision-making, and real-time monitoring and early warning. The intelligent business opportunity business processing optimization method provided by the present invention belongs to the field of artificial intelligence technology, and aims to solve the technical problems in the existing technology such as incomplete business opportunity feature extraction, insufficient optimization of resource allocation, and insufficient timeliness of execution monitoring. Summary of the invention

[0005] The purpose of this section is to summarize some aspects of embodiments of the present invention and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section and the specification abstract and the invention title of this application to avoid blurring the purpose of this section, the specification abstract and the invention title, and such simplifications or omissions cannot be used to limit the scope of the present invention.

[0006] In view of the above existing problems, the present invention is proposed.

[0007] Therefore, the present invention provides an intelligent business opportunity service processing optimization method and system, which can solve the problems mentioned in the background technology.

[0008] In order to solve the above technical problems, the present invention provides the following technical solutions: In a first aspect, the present invention provides an intelligent business opportunity service processing optimization method, which includes collecting historical business opportunity data and performing multi-dimensional feature extraction to obtain business opportunity feature data; Building a time series prediction network based on the attention mechanism, performing data processing on the business opportunity feature data, and obtaining business opportunity conversion probability data; Based on the business opportunity conversion probability data, combined with the sales team capability data and resource utilization data, a business opportunity allocation plan is generated through a reinforcement learning algorithm; The execution status of the business opportunity allocation plan is monitored, and when it is detected that the status indicator deviates from the preset threshold, warning information and intervention suggestions are output.

[0009] As a preferred solution of the intelligent business opportunity service processing optimization method of the present invention, wherein: the business opportunity feature data obtained includes: Divide the historical business opportunity data into multiple time windows according to the time series, perform data cleaning on the business opportunity data in each time window, and generate standardized business opportunity data; Extracting customer portrait features from the standardized business opportunity data, the customer portrait features including historical transaction amount, transaction frequency, industry type and enterprise size, and calculating the timeliness weight of the customer portrait features based on a time decay function to generate weighted customer portrait features; Collecting market trend characteristics and competition situation characteristics corresponding to the standardized business opportunity data, wherein the market trend characteristics include industry growth rate, market capacity and policy orientation index, and the competition situation characteristics include the number of competing products, market share and price index; The weighted customer portrait features, the market trend features and the competition situation features are input into a feature fusion network to generate a business opportunity feature vector, and the business opportunity feature vector is subjected to dimensionality reduction processing to obtain business opportunity feature data.

[0010] As a preferred solution of the intelligent business opportunity service processing optimization method of the present invention, wherein: the business opportunity conversion probability data obtained includes: Construct an input sequence from the business opportunity feature data in chronological order, perform sliding window segmentation on the input sequence, and generate feature sequence samples; Construct a multi-head attention layer, wherein the multi-head attention layer includes K attention heads, each of which independently calculates the attention weight of each time step in the feature sequence sample, and performs weighted calculation on each time step feature in the feature sequence sample and the corresponding attention weight to generate an attention weighted feature; Inputting the attention weighted features into a bidirectional long short-term memory network, extracting the temporal dependency of the feature sequence samples, and outputting temporal features; A neural network classifier is trained based on historical business opportunity conversion labels, the neural network classifier uses cross entropy as a loss function, the time series features are input into the neural network classifier, and business opportunity conversion probability data is calculated.

[0011] As a preferred solution of the intelligent business opportunity service processing optimization method of the present invention, wherein: the construction of the time series prediction network includes: Sort the business opportunity feature data by timestamp, and for each business opportunity, retain the complete feature vector and corresponding time information to form an initial time series data set; Use sliding window technology to segment time series data, set a fixed-size time window and sliding step size, and segment the feature sequence; The feature sequence in each sliding window is sampled to generate feature sequence samples of equal length. If the feature sequence length is less than the standard length, the time step filling strategy is used to fill it. If the feature sequence length is greater than the standard length, it is truncated.

[0012] As a preferred solution of the intelligent business opportunity service processing optimization method of the present invention, wherein: the generating business opportunity allocation plan includes: Collect historical performance data, professional field data and current load data of sales team members, build a sales team capability matrix, and standardize the business opportunity conversion probability data and the sales team capability matrix to generate a state vector; Construct a reinforcement learning model based on a deep Q network, taking the state vector as input and the business opportunity allocation decision as the action space, and setting a reward function that comprehensively calculates business opportunity conversion benefits, resource balance, and team matching; An experience replay mechanism is used to store historical allocation decision data, wherein the historical allocation decision data includes state transition samples, action samples, and reward samples, and the deep Q network is trained based on the historical allocation decision data to update network parameters; The current state vector is input into the trained deep Q network, and a business opportunity allocation plan is output.

[0013] As a preferred solution of the intelligent business opportunity service processing optimization method of the present invention, wherein: the construction of the reinforcement learning model based on the deep Q network includes: Design a deep Q network structure, where the number of nodes in the input layer corresponds to the dimension of the state vector, the number of nodes in the output layer corresponds to the dimension of the action space, and the intermediate layer adopts a residual connection structure; Define a discrete action space and construct all possible combinations of allocating business opportunities to sales team members as an action set; Design a composite reward function, the calculation formula is as follows: ; in, Represents the comprehensive reward value, represents the conversion profit function, represents the resource balancing function, represents the professional matching function, , , Indicates the reward item weight; The conversion revenue function calculation formula is as follows: ; in, Indicates the number of business opportunities to be allocated. represents the expected value of the ith business opportunity, represents the conversion probability of the i-th business opportunity, Indicates the expected transaction cycle. represents the time attenuation coefficient; The resource balancing function calculation formula is as follows: ; in, Indicates the number of sales team members, represents the normalized workload of the jth member, Indicates the average load level of the team; The calculation formula of professional matching function is as follows: ; in, Indicates the number of professional dimensions, represents the kth professional dimension weight, represents the demand vector of the i-th business opportunity in the k-th professional dimension, represents the capability vector of the assigned sales member in the kth professional dimension, represents the vector similarity calculation function, Indicates the assignment object of the i-th business opportunity.

[0014] As a preferred solution of the intelligent business opportunity service processing optimization method of the present invention, the output of warning information and intervention suggestions includes: Collect follow-up frequency data, customer feedback data and transaction progress data during the business opportunity processing process, generate execution status data, and build a multi-dimensional monitoring indicator system to map the execution status data into response time indicators, interaction quality indicators and progress achievement indicators; The threshold intervals of the response time indicator, the interaction quality indicator and the progress achievement indicator are set based on historical success cases, and a sliding time window is used to calculate the change trend of the execution status data to generate a trend feature sequence; Input the trend feature sequence into an anomaly detection model based on an isolation forest algorithm to identify an abnormal pattern, and when an abnormal pattern is detected, analyze the cause of the abnormality based on a decision tree model, wherein the input variables of the decision tree model include the execution status data and the business opportunity allocation plan; Generate targeted intervention suggestions based on the abnormal cause, and the intervention suggestions include resource allocation plans and processing strategy optimization plans.

[0015] In a second aspect, the present invention provides an intelligent business opportunity processing optimization system, which includes: a business opportunity feature extraction module, a conversion probability prediction module, an intelligent allocation decision module, and an execution monitoring and early warning module; The business opportunity feature extraction module is used to collect historical business opportunity data and perform multi-dimensional feature extraction to obtain business opportunity feature data; The conversion probability prediction module is used to construct a time series prediction network based on the attention mechanism, perform data processing on the business opportunity feature data, and obtain business opportunity conversion probability data; The intelligent allocation decision module is used to generate a business opportunity allocation plan through a reinforcement learning algorithm based on the business opportunity conversion probability data, combined with the sales team capability data and resource utilization data; The execution monitoring and early warning module is used to monitor the execution status of the business opportunity allocation plan, and output early warning information and intervention suggestions when it is detected that the status indicator deviates from the preset threshold.

[0016] In a third aspect, the present invention provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, wherein: when the processor executes the computer program, the steps of the intelligent business opportunity business processing optimization method are implemented.

[0017] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: the computer program, when executed by a processor, implements the steps of an intelligent business opportunity service processing optimization method.

[0018] Compared with the prior art, the beneficial effects of the present invention are as follows: the perception of the development laws of business opportunities is enhanced through a time series prediction network based on an attention mechanism, and the accuracy of conversion probability prediction is significantly improved; dynamic optimization allocation of sales resources is realized with the help of a deep reinforcement learning algorithm, and resource utilization efficiency is effectively improved; at the same time, anomaly detection and decision tree analysis based on isolation forests are used to realize intelligent monitoring and timely intervention of the execution process; the organic combination with the above-mentioned technical solutions constructs a closed-loop intelligent business opportunity management system, which not only significantly improves the automation and intelligence level of business opportunity management, but also enhances the system's adaptability through the innovative application of algorithm models, and ultimately achieves a dual improvement in business opportunity processing efficiency and conversion rate, providing reliable technical support for the digital transformation of enterprise sales management. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0020] Figure 1 A method flow chart of an intelligent business opportunity service processing optimization method and system provided by one embodiment of the present invention; Figure 2 An internal structure diagram of a computer device of an intelligent business opportunity service processing optimization method and system provided by an embodiment of the present invention; Figure 3 A system module diagram of an intelligent business opportunity service processing optimization method and system provided for one embodiment of the present invention. DETAILED DESCRIPTION

[0021] In order to make the above-mentioned purposes, features and advantages of the present invention more understandable, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0022] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0023] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0024] Example 1, reference Figure 1-Figure 2 , which is the first embodiment of the present invention, and provides an intelligent business opportunity service processing optimization method, comprising: Before describing the embodiments of the present application in detail, some related concepts are first explained for the sake of clarity.

[0025] Attention Mechanism: This is an important mechanism in deep learning. Its core idea is to allow the model to automatically identify and focus on important parts of information when processing input sequences. In business opportunity forecasting, the attention mechanism can automatically identify and highlight key time nodes and important features in the development process of business opportunities, thereby improving forecast accuracy.

[0026] Isolation Forest: This is an anomaly detection algorithm. Its basic principle is to isolate abnormal points by randomly constructing decision trees. This algorithm believes that abnormal data points are easier to isolate than normal data points. In this invention, it is used to identify abnormal states in the process of business opportunity processing. It has the characteristics of high computational efficiency and strong ability to process high-dimensional data.

[0027] Bidirectional LSTM: This is a special recurrent neural network structure that contains two LSTM layers, forward and backward, and can use both past and future context information. In this invention, this network is used to capture long-term dependencies in the process of business opportunity development and improve the effect of time series feature extraction.

[0028] Experience Replay: This is a key technology in reinforcement learning that improves learning efficiency by storing and reusing historical decision-making experience. In this invention, this mechanism is used to store historical business opportunity allocation decision data to help the deep Q network better learn the optimal allocation strategy.

[0029] This application provides a method that can effectively solve the above-mentioned problems. Next, we will combine multiple embodiments to explain in detail how to implement the intelligent business opportunity service processing optimization method; Figure 1 A method flow chart of an intelligent business opportunity business processing optimization method and system is shown, including: S1: Collect historical business opportunity data and perform multi-dimensional feature extraction to obtain business opportunity feature data; Furthermore, the historical business opportunity data is divided into multiple time windows according to the time series, and the business opportunity data in each time window is cleaned to generate standardized business opportunity data; Extracting customer portrait features from the standardized business opportunity data, the customer portrait features including historical transaction amount, transaction frequency, industry type and enterprise size, and calculating the timeliness weight of the customer portrait features based on a time decay function to generate weighted customer portrait features; Collecting market trend characteristics and competition situation characteristics corresponding to the standardized business opportunity data, wherein the market trend characteristics include industry growth rate, market capacity and policy orientation index, and the competition situation characteristics include the number of competing products, market share and price index; The weighted customer portrait features, the market trend features and the competition situation features are input into a feature fusion network to generate a business opportunity feature vector, and the business opportunity feature vector is subjected to dimensionality reduction processing to obtain business opportunity feature data.

[0030] Furthermore, generating standardized business opportunity data includes: Divide the historical business opportunity data into time windows according to a fixed time span; it should be noted that the present invention sets the fixed time span to 3 months. This division method ensures the continuity and integrity of the data. Each time window contains all business opportunity records within its interval. The system will assign a unique identifier to each time window and record the start and end time of the window; Perform missing value processing on the data in each time window, and use different processing strategies for different types of features; the processing strategies include: for numerical features (such as sales amount, number of contacts, etc.), use the median in the time window for filling, which is insensitive to outliers; for categorical features (such as industry type, customer level, etc.), use the category (mode) with the highest frequency in the window for filling, and record the missing rate and filling value of each feature for subsequent data quality assessment; Outlier processing is performed on the data in each time window, and the 3σ principle in statistics is used to identify outliers. The calculation formula is as follows: ; in, represents the eigenvalue after outlier processing, represents the original eigenvalue, represents the feature mean, represents the characteristic standard deviation; The minimum-maximum normalization method is used to uniformly map all numerical features to the [0,1] interval.

[0031] Furthermore, the features of customer portraits are extracted including: The system extracts key behavioral characteristics of customers from historical transaction records, including historical transaction amount, transaction frequency, industry type, and enterprise size; The historical transaction amount includes the total transaction amount and the average transaction amount in the past 12 months; The transaction frequency includes calculating the customer's purchase cycle and purchase frequency; The industry type includes using a hierarchical coding method to convert industry information into numerical features; The enterprise scale includes comprehensive consideration of factors such as registered capital and number of employees to construct an enterprise scale indicator. It should be noted that in order to ensure the timeliness of the features, the system has designed a weighting mechanism based on time decay. The farther the historical data is from the current time, the lower its weight, which reflects the timeliness characteristics of business behavior. The decay weight is calculated through an exponential function to ensure the real-time and effectiveness of the features.

[0032] Furthermore, the business opportunity feature data includes: A multi-layer neural network structure is used for feature fusion, which deeply integrates customer portrait features, market trend features, and competitive situation features to generate a unified feature representation. It should be noted that the network uses the ReLU activation function to ensure the nonlinear expression capability of feature extraction. The principal component analysis method is used to reduce the dimension of the fused features, retaining the principal components with a cumulative variance contribution rate of 95%. This processing not only retains the main information of the data, but also reduces the complexity of subsequent modeling.

[0033] S2: construct a time series prediction network based on the attention mechanism, process the business opportunity feature data, and obtain business opportunity conversion probability data; Furthermore, the business opportunity conversion probability data includes: Construct an input sequence from the business opportunity feature data in chronological order, perform sliding window segmentation on the input sequence, and generate feature sequence samples; Construct a multi-head attention layer, wherein the multi-head attention layer includes K attention heads, each of which independently calculates the attention weight of each time step in the feature sequence sample, and performs weighted calculation on each time step feature in the feature sequence sample and the corresponding attention weight to generate an attention weighted feature; Inputting the attention weighted features into a bidirectional long short-term memory network, extracting the temporal dependency of the feature sequence samples, and outputting temporal features; A neural network classifier is trained based on historical business opportunity conversion labels, the neural network classifier uses cross entropy as a loss function, the time series features are input into the neural network classifier, and business opportunity conversion probability data is calculated.

[0034] Furthermore, constructing a time series prediction network includes: Sort the business opportunity feature data by timestamp to ensure the time continuity of the data. For each business opportunity, retain its complete feature vector and corresponding time information to form the initial time series data set. The sliding window technology is used to segment the time series data, and a fixed-size time window and sliding step are set to segment the feature sequence. It should be noted that each window contains a continuous time series data, and there is overlap between adjacent windows. This overlapping design can capture the gradual change process of the feature. The feature sequence in each sliding window is sampled to generate feature sequence samples of equal length. If the feature sequence length is less than the standard length, the time step filling strategy is used to fill it. If the feature sequence length is greater than the standard length, it is truncated. This ensures that all sequence samples have the same time step length, which is convenient for subsequent batch processing.

[0035] Furthermore, constructing a multi-head attention layer includes: K independent attention heads are designed. Each attention head is responsible for learning temporal associations from different feature representation spaces. The attention calculation adopts the scaled dot product attention mechanism. The calculation formula is as follows: ; Among them, Q represents the query matrix, K represents the key matrix, and V represents the value matrix. represents the feature dimension, and softmax represents the normalization function; Each attention head independently calculates the attention score of each time step in the feature sequence and performs weighted aggregation. The outputs of K attention heads are concatenated and the final attention weighted features are generated through linear transformation.

[0036] Furthermore, extracting the temporal dependency of the characteristic sequence samples includes: Construct a bidirectional LSTM network structure, including forward and backward LSTM layers. The forward LSTM layer extracts time series features from the past to the present, and the backward LSTM layer extracts time series features from the present to the past. This bidirectional structure can make full use of context information. The attention-weighted features are input into the bidirectional LSTM network in time steps. The hidden state of each time step contains the timing information before and after that moment. The hidden states in the two directions are fused to generate a feature representation containing complete temporal dependencies.

[0037] Furthermore, training a neural network classifier includes, Design a multi-layer neural network classifier, including a fully connected layer, a batch normalization layer, and a dropout layer. The input layer dimension matches the time series feature dimension. The output layer uses a sigmoid activation function to output the business opportunity conversion probability. The cross entropy loss function is used for model training. The loss function is as follows: ; in, represents the loss function value, represents the total number of samples, represents the true label of the i-th sample, represents the predicted probability of the i-th sample; Use the Adam optimizer for parameter optimization, adopt the early stopping strategy to avoid overfitting, monitor the model performance on the validation set, select the optimal model parameters, and record the loss value, accuracy and other indicators during the training process for model effect evaluation and tuning.

[0038] S3: Based on the business opportunity conversion probability data, combined with the sales team capability data and resource utilization data, a business opportunity allocation plan is generated through a reinforcement learning algorithm; Furthermore, generating a business opportunity allocation plan includes: Collect historical performance data, professional field data and current load data of sales team members, build a sales team capability matrix, and standardize the business opportunity conversion probability data and the sales team capability matrix to generate a state vector; Construct a reinforcement learning model based on a deep Q network, taking the state vector as input and the business opportunity allocation decision as the action space, and setting a reward function that comprehensively calculates business opportunity conversion benefits, resource balance, and team matching; An experience replay mechanism is used to store historical allocation decision data, wherein the historical allocation decision data includes state transition samples, action samples, and reward samples, and the deep Q network is trained based on the historical allocation decision data to update network parameters; The current state vector is input into the trained deep Q network, and a business opportunity allocation plan is output.

[0039] Furthermore, building a sales team capability matrix includes: By collecting the performance data of each sales team member in the past 12 months, including the total transaction amount, transaction volume, and average order value distribution, the transaction status of each member in different industries and with different sizes of customers is classified and counted to form a performance vector; Establish a professional database for sales team members, record each member's industry experience, product expertise, and technical certification level, and convert it into a professional capability vector through quantitative scoring standards; Count the current load of sales team members, including the number of business opportunities being followed up, the number of scheduled visits, and the frequency of customer communication, and integrate them into a load vector; The sales team capability matrix is ​​constructed by combining the performance vector, professional capability vector and load vector. The calculation formula is as follows: ; in, Represents the sales team capability matrix, represents the performance vector, represents the professional capability vector, represents the load vector.

[0040] Furthermore, generating the state vector includes, Performing distribution mapping on the business opportunity conversion probability data to generate a business opportunity feature vector; Regularizing the sales team capability matrix to generate a standardized capability matrix; The opportunity feature vector is combined with the standardized capability matrix and mapped into a state vector: ; in, represents the state vector at time t, represents the business opportunity feature vector, represents the standardized capability matrix, Represents a tensor interoperation operator.

[0041] Furthermore, the construction of a reinforcement learning model based on a deep Q network includes: Design a deep Q network structure, where the number of nodes in the input layer corresponds to the dimension of the state vector, the number of nodes in the output layer corresponds to the dimension of the action space, and the intermediate layer adopts a residual connection structure; Define a discrete action space and construct all possible combinations of allocating business opportunities to sales team members as an action set; Design a compound reward function, the calculation formula is as follows: ; in, Represents the comprehensive reward value, represents the conversion profit function, represents the resource balancing function, represents the professional matching function, , , Indicates the reward item weight.

[0042] Specifically, the conversion revenue function calculation formula is as follows: ; in, Indicates the number of business opportunities to be allocated. represents the expected value of the ith business opportunity, represents the conversion probability of the i-th business opportunity, Indicates the expected transaction cycle. represents the time attenuation coefficient; The resource balancing function calculation formula is as follows: ; in, Indicates the number of sales team members, represents the normalized workload of the jth member, Indicates the average load level of the team; The calculation formula of professional matching function is as follows: ; in, Indicates the number of professional dimensions, represents the kth professional dimension weight, represents the demand vector of the i-th business opportunity in the k-th professional dimension, represents the capability vector of the assigned sales member in the kth professional dimension, represents the vector similarity calculation function, Indicates the assignment object of the i-th business opportunity.

[0043] Further, training the deep Q network includes, Establish an experience replay pool to store historical allocation decision data, including state transition samples, action samples, and reward samples; Randomly extract sample batches from the experience replay pool, calculate the temporal difference error, and update the deep Q network parameters; Input the current state vector into the trained deep Q network, and select the action with the largest Q value as the business opportunity allocation plan; Execute the business opportunity allocation plan and record new decision data into the experience replay pool for subsequent model training.

[0044] S4: monitoring the execution status of the business opportunity allocation plan, and outputting warning information and intervention suggestions when it is detected that the status indicator deviates from the preset threshold.

[0045] Further, the output of warning information and intervention suggestions includes: Collect follow-up frequency data, customer feedback data and transaction progress data during the business opportunity processing process, generate execution status data, and build a multi-dimensional monitoring indicator system to map the execution status data into response time indicators, interaction quality indicators and progress achievement indicators; The threshold intervals of the response time indicator, the interaction quality indicator and the progress achievement indicator are set based on historical success cases, and a sliding time window is used to calculate the change trend of the execution status data to generate a trend feature sequence; Input the trend feature sequence into an anomaly detection model based on an isolation forest algorithm to identify an abnormal pattern, and when an abnormal pattern is detected, analyze the cause of the abnormality based on a decision tree model, wherein the input variables of the decision tree model include the execution status data and the business opportunity allocation plan; Generate targeted intervention suggestions based on the abnormal cause, and the intervention suggestions include resource allocation plans and processing strategy optimization plans.

[0046] Example 2, reference Figure 3 ,This embodiment also provides an intelligent business opportunity business processing optimization system, including: a business opportunity feature extraction module, a conversion probability prediction module, an intelligent allocation decision module, and an execution monitoring and early warning module; The business opportunity feature extraction module is used to collect historical business opportunity data and perform multi-dimensional feature extraction to obtain business opportunity feature data; The conversion probability prediction module is used to construct a time series prediction network based on the attention mechanism, perform data processing on the business opportunity feature data, and obtain business opportunity conversion probability data; The intelligent allocation decision module is used to generate a business opportunity allocation plan through a reinforcement learning algorithm based on the business opportunity conversion probability data, combined with the sales team capability data and resource utilization data; The execution monitoring and early warning module is used to monitor the execution status of the business opportunity allocation plan, and output early warning information and intervention suggestions when it is detected that the status indicator deviates from the preset threshold.

[0047] This embodiment also provides a computer device, which may be a terminal, and its internal structure diagram may be as shown in FIG. Figure 2As shown. The computer device includes a processor, a memory, a communication interface, a display screen and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. When the computer program is executed by the processor, a method for optimizing intelligent business opportunity business processing is implemented. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad set on the computer device housing, or an external keyboard, touchpad or mouse, etc.

[0048] This embodiment also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented: collecting historical business opportunity data, and performing multi-dimensional feature extraction to obtain business opportunity feature data; Building a time series prediction network based on the attention mechanism, performing data processing on the business opportunity feature data, and obtaining business opportunity conversion probability data; Based on the business opportunity conversion probability data, combined with the sales team capability data and resource utilization data, a business opportunity allocation plan is generated through a reinforcement learning algorithm; The execution status of the business opportunity allocation plan is monitored, and when it is detected that the status indicator deviates from the preset threshold, warning information and intervention suggestions are output.

[0049] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

[0050] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes. The schemes in the embodiments of the present application may be implemented in various computer languages, for example, object-oriented programming language Java and literal scripting language JavaScript, etc.

[0051] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0052] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0053] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0054] Although the preferred embodiments of the present application have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications falling within the scope of the present application.

[0055] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is also intended to include these modifications and variations.

Claims

1. An intelligent business opportunity service processing optimization method, characterized in that: Including collecting historical business opportunity data and performing multi-dimensional feature extraction to obtain business opportunity feature data; Building a time series prediction network based on the attention mechanism, performing data processing on the business opportunity feature data, and obtaining business opportunity conversion probability data; Based on the business opportunity conversion probability data, combined with the sales team capability data and resource utilization data, a business opportunity allocation plan is generated through a reinforcement learning algorithm; The execution status of the business opportunity allocation plan is monitored, and when it is detected that the status indicator deviates from the preset threshold, warning information and intervention suggestions are output.

2. The intelligent business opportunity service processing optimization method according to claim 1, characterized in that: The business opportunity characteristic data obtained includes: Divide the historical business opportunity data into multiple time windows according to the time series, perform data cleaning on the business opportunity data in each time window, and generate standardized business opportunity data; Extracting customer portrait features from the standardized business opportunity data, the customer portrait features including historical transaction amount, transaction frequency, industry type and enterprise size, and calculating the timeliness weight of the customer portrait features based on a time decay function to generate weighted customer portrait features; Collecting market trend characteristics and competition situation characteristics corresponding to the standardized business opportunity data, wherein the market trend characteristics include industry growth rate, market capacity and policy orientation index, and the competition situation characteristics include the number of competing products, market share and price index; The weighted customer portrait features, the market trend features and the competition situation features are input into a feature fusion network to generate a business opportunity feature vector, and the business opportunity feature vector is subjected to dimensionality reduction processing to obtain business opportunity feature data.

3. The intelligent business opportunity service processing optimization method according to claim 2, characterized in that: The business opportunity conversion probability data obtained includes: Construct an input sequence from the business opportunity feature data in chronological order, perform sliding window segmentation on the input sequence, and generate feature sequence samples; Constructing a multi-head attention layer, the multi-head attention layer includes K attention heads, each of which independently calculates the attention weight of each time step in the feature sequence sample, and performs weighted calculation on each time step feature in the feature sequence sample and the corresponding attention weight to generate an attention weighted feature; Inputting the attention weighted features into a bidirectional long short-term memory network, extracting the temporal dependency of the feature sequence samples, and outputting temporal features; A neural network classifier is trained based on historical business opportunity conversion labels. The neural network classifier uses cross entropy as a loss function. The time series features are input into the neural network classifier to calculate business opportunity conversion probability data.

4. The intelligent business opportunity service processing optimization method according to claim 3, characterized in that: The construction of the time series prediction network includes: Sort the business opportunity feature data by timestamp, and for each business opportunity, retain the complete feature vector and corresponding time information to form an initial time series data set; Use sliding window technology to segment time series data, set a fixed-size time window and sliding step size, and segment the feature sequence; The feature sequence in each sliding window is sampled to generate feature sequence samples of equal length. If the feature sequence length is less than the standard length, the time step filling strategy is used to fill it. If the feature sequence length is greater than the standard length, it is truncated.

5. The intelligent business opportunity service processing optimization method according to claim 4, characterized in that: The generating business opportunity allocation plan includes: Collect historical performance data, professional field data and current load data of sales team members, build a sales team capability matrix, and standardize the business opportunity conversion probability data and the sales team capability matrix to generate a state vector; Construct a reinforcement learning model based on a deep Q network, taking the state vector as input and the business opportunity allocation decision as the action space, and setting a reward function that comprehensively calculates business opportunity conversion benefits, resource balance, and team matching; An experience replay mechanism is used to store historical allocation decision data, wherein the historical allocation decision data includes state transition samples, action samples, and reward samples, and the deep Q network is trained based on the historical allocation decision data to update network parameters; The current state vector is input into the trained deep Q network, and a business opportunity allocation plan is output.

6. The intelligent business opportunity service processing optimization method according to claim 5, characterized in that: The construction of a reinforcement learning model based on a deep Q network includes: Design a deep Q network structure, where the number of nodes in the input layer corresponds to the dimension of the state vector, the number of nodes in the output layer corresponds to the dimension of the action space, and the intermediate layer adopts a residual connection structure; Define a discrete action space and construct all possible combinations of allocating business opportunities to sales team members as an action set; Design a composite reward function, the calculation formula is as follows: ； in, Represents the comprehensive reward value, represents the conversion profit function, represents the resource balancing function, represents the professional matching function, , , Indicates the weight of reward items; The conversion revenue function calculation formula is as follows: ； in, Indicates the number of business opportunities to be allocated. represents the expected value of the ith business opportunity, represents the conversion probability of the i-th business opportunity, Indicates the expected transaction cycle. represents the time attenuation coefficient; The resource balancing function calculation formula is as follows: ； in, Indicates the number of sales team members, represents the normalized workload of the jth member, Indicates the average load level of the team; The calculation formula of professional matching function is as follows: ； in, Indicates the number of professional dimensions, represents the kth professional dimension weight, represents the demand vector of the i-th business opportunity in the k-th professional dimension, represents the capability vector of the assigned sales member in the kth professional dimension, represents the vector similarity calculation function, Indicates the assignment object of the i-th business opportunity.

7. The intelligent business opportunity service processing optimization method according to claim 6, characterized in that: The output of warning information and intervention suggestions includes: Collect follow-up frequency data, customer feedback data and transaction progress data during the business opportunity processing process, generate execution status data, and build a multi-dimensional monitoring indicator system to map the execution status data into response time indicators, interaction quality indicators and progress achievement indicators; The threshold intervals of the response time indicator, the interaction quality indicator and the progress achievement indicator are set based on historical success cases, and a sliding time window is used to calculate the change trend of the execution status data to generate a trend feature sequence; Input the trend feature sequence into an anomaly detection model based on an isolation forest algorithm to identify an abnormal pattern, and when an abnormal pattern is detected, analyze the cause of the abnormality based on a decision tree model, wherein the input variables of the decision tree model include the execution status data and the business opportunity allocation plan; Generate targeted intervention suggestions based on the abnormal cause, and the intervention suggestions include resource allocation plans and processing strategy optimization plans.

8. An intelligent business opportunity service processing optimization system, based on the intelligent business opportunity service processing optimization method according to any one of claims 1 to 7, characterized in that: It includes business opportunity feature extraction module, conversion probability prediction module, intelligent allocation decision module and execution monitoring and early warning module; The business opportunity feature extraction module is used to collect historical business opportunity data and perform multi-dimensional feature extraction to obtain business opportunity feature data; The conversion probability prediction module is used to construct a time series prediction network based on the attention mechanism, perform data processing on the business opportunity feature data, and obtain business opportunity conversion probability data; The intelligent allocation decision module is used to generate a business opportunity allocation plan through a reinforcement learning algorithm based on the business opportunity conversion probability data, combined with the sales team capability data and resource utilization data; The execution monitoring and early warning module is used to monitor the execution status of the business opportunity allocation plan, and output early warning information and intervention suggestions when it is detected that the status indicator deviates from the preset threshold.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the intelligent business opportunity service processing optimization method described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the intelligent business opportunity service processing optimization method described in any one of claims 1 to 7 are implemented.

Images