Customer resource management system and method based on large model

Abstract

The invention relates to a customer resource management system and method based on a large model. The method comprises the following steps: carrying out semantic analysis on historical sales data of sales personnel and new customer resources by adopting a natural language processing technology based on deep learning, mining professional experience information of the sales personnel and potential demand information of the customer resources, and carrying out fine-grained correlation matching analysis on the professional experience information and the potential demand information; in this way, whether the new customer resources are matched with the ability and style of the salesman or not is intelligently judged, and therefore accurate pushing and optimal management of the customer resources are achieved. In this way, the efficiency and accuracy of customer resource allocation can be remarkably improved, the invalid communication cost in the sales process is reduced, the customer satisfaction is improved, and enterprises are assisted in achieving continuous increase of sales performance.

Description

Technical Field

[0001] The present application relates to the technical field of customer resource management, and in particular, to a customer resource management system and method based on a large model. Background Art

[0002] In the current business environment, customer resource management directly affects sales performance and customer satisfaction. Especially for sales-driven companies, customer resource management has become a core strategy for companies to achieve sustainable development and competitive advantage.

[0003] Customer resource management not only involves the collection and analysis of customer data, but also how to effectively use customer information to optimize business processes, improve customer satisfaction, and achieve sales targets. However, traditional customer resource management methods often rely on manual experience to allocate customer resources, which is not only inefficient but also easily affected by subjective factors and difficult to adapt to the rapidly changing market environment. Therefore, an optimized customer resource management system and method are expected. Summary of the invention

[0004] This summary is provided to introduce concepts in a brief form that will be described in detail in the detailed description below. This summary is not intended to identify key features or essential features of the claimed technical solution, nor is it intended to limit the scope of the claimed technical solution.

[0005] In a first aspect, the present application provides a customer resource management system based on a large model, the system comprising:

[0006] The customer resource information acquisition module is used to obtain the text description of new customer resources;

[0007] A salesperson historical sales data acquisition module is used to acquire the historical sales data of candidate salespersons;

[0008] A semantic coding module, used for semantically coding the text description of the new customer resource and each sales data in the historical sales data of the candidate salesperson, respectively, to obtain a set of new customer resource semantic coding feature vectors and sales data semantic coding feature vectors;

[0009] A semantic association strengthening module, used for performing semantic association strengthening on the set of sales data semantic coding feature vectors to obtain a set of strengthened sales data semantic coding feature vectors;

[0010] The resource push module is used to determine whether to push the new customer resources to the candidate salesperson based on the fine-grained association matching between the semantic coding feature vector of the new customer resources and the set of semantic coding feature vectors of the enhanced sales data.

[0011] Optionally, the semantic association enhancement module is used to: perform a semantic association enhancement module based on feature energy level measurement association radiation on the set of sales data semantic encoding feature vectors to obtain the set of enhanced sales data semantic encoding feature vectors.

[0012] Optionally, the semantic association enhancement module includes: a characteristic energy level coefficient calculation unit, used to calculate the characteristic energy level coefficients of each sales data semantic coding feature vector in the set of sales data semantic coding feature vectors to obtain a set of characteristic energy level coefficients; a one-dimensional arrangement unit, used to arrange the set of characteristic energy level coefficients into a characteristic energy level coefficient input vector; a characteristic energy level radiation association encoding unit, used to perform one-dimensional convolution encoding on the characteristic energy level coefficient input vector to obtain a characteristic energy level radiation association vector; a probabilistic mapping unit, used to perform point convolution encoding and probabilistic mapping based on the Sigmoid function on the characteristic energy level radiation association vector to obtain a characteristic energy level radiation association weight vector; a semantic enhancement unit, used to perform weighted processing and residual processing on the set of sales data semantic coding feature vectors using the characteristic energy level radiation association weight vector as a weight vector to obtain the set of enhanced sales data semantic coding feature vectors.

[0013] Optionally, the feature energy level coefficient calculation unit is used to: calculate the weighted sum of the variance of the sales data semantic encoding feature vector and the bias term to obtain a first intermediate term, and calculate the square of the difference between the maximum value and the mean of the sales data semantic encoding feature vector, the variance of the sales data semantic encoding feature vector, and the bias term to obtain a second intermediate term; calculate the first intermediate term divided by the second intermediate term to obtain the feature energy level coefficient.

[0014] Optionally, the resource push module includes: a fine-grained association matching analysis unit, used to perform fine-grained association matching analysis on the set of the new customer resource semantic encoding feature vector and the enhanced sales data semantic encoding feature vector to obtain a query response adaptive matching result representation matrix; inputting the query response adaptive matching result representation matrix into a classifier-based customer resource management result generator to obtain a management result, and the management result is used to indicate whether to push new customer resources to alternative sales personnel.

[0015] Optionally, the fine-grained association matching analysis unit is used to: use the new customer resource semantic encoding feature vector as a query feature vector, and input the query feature vector and a set of the enhanced sales data semantic encoding feature vector into a feature query response adaptive encoder based on fine-grained decoupling to obtain the query response adaptive matching result representation matrix.

[0016] Optionally, the fine-grained association matching analysis unit includes: a local semantic association reinforcement subunit, which is used to perform local semantic association reinforcement on each enhanced sales data semantic encoding feature vector in the set of enhanced sales data semantic encoding feature vectors to obtain a set of sales data fine-grained decoupled local semantic association feature vectors; a covariance calculation subunit, which is used to calculate the covariance matrix between the query feature vector and each sales data fine-grained decoupled local semantic association feature vector in the set of sales data fine-grained decoupled local semantic association feature vectors to obtain a set of query response covariance matrices; a matrix trace calculation subunit, which is used to calculate the matrix trace of each query response covariance matrix in the set of query response covariance matrices to obtain a set of query response trace measurement factors; a weighting subunit, which is used to input the set of query response trace measurement factors into a Sigmoid activation function to obtain a set of query response trace attention weights; an attention application subunit, which is used to calculate the weighted sum of the set of query response covariance matrices using each query response trace attention weight in the set of query response trace attention weights as a weight to obtain the query response adaptive matching result representation matrix.

[0017] Optionally, the local semantic association reinforcement subunit is used to: perform fine-grained decoupling of the enhanced sales data semantic encoding feature vector to obtain a sequence of enhanced sales data local semantic encoding feature vectors; input the sequence of enhanced sales data local semantic encoding feature vectors into a sequence semantic reinforcement encoder based on a Bi-LSTM model to obtain the fine-grained decoupled local semantic association feature vector of the sales data.

[0018] In a second aspect, the present application provides a customer resource management method based on a large model, the method comprising:

[0019] Get the text description of the new customer resource,

[0020] Obtain historical sales data of candidate sales personnel;

[0021] Semantically encoding the text description of the new customer resource and each sales data in the historical sales data of the candidate salesperson to obtain a set of new customer resource semantic encoding feature vectors and sales data semantic encoding feature vectors;

[0022] Performing semantic association enhancement on the set of sales data semantic encoding feature vectors to obtain a set of enhanced sales data semantic encoding feature vectors;

[0023] Based on the fine-grained association matching between the new customer resource semantic encoding feature vector and the set of enhanced sales data semantic encoding feature vectors, it is determined whether to push the new customer resource to the candidate salesperson.

[0024] Optionally, semantic association enhancement is performed on the set of sales data semantic coding feature vectors to obtain a set of enhanced sales data semantic coding feature vectors, including: performing a semantic association enhancement module based on feature energy level metric association radiation on the set of sales data semantic coding feature vectors to obtain the set of enhanced sales data semantic coding feature vectors.

[0025] The above technical solution uses natural language processing technology based on deep learning to perform semantic analysis on the sales staff's historical sales data and new customer resources, dig out the sales staff's professional experience information and the potential demand information of customer resources, and conduct fine-grained correlation matching analysis on the two to intelligently determine whether the new customer resources match the sales staff's capabilities and style, thereby achieving accurate push and optimized management of customer resources. In this way, the efficiency and accuracy of customer resource allocation can be significantly improved, the cost of ineffective communication in the sales process can be reduced, customer satisfaction can be improved, and enterprises can achieve sustained growth in sales performance.

[0026] Other features and advantages of the present application will be described in detail in the subsequent detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] The above and other features, advantages and aspects of the embodiments of the present application will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and the originals and elements are not necessarily drawn to scale. In the drawings:

[0028] Figure 1 It is a block diagram of a customer resource management system based on a big model according to an exemplary embodiment.

[0029] Figure 2 The figure is a flow chart of a customer resource management method based on a big model according to an exemplary embodiment.

[0030] Figure 3 It is a block diagram of an electronic device according to an exemplary embodiment.

[0031] Figure 4 It is an application scenario diagram of a customer resource management system based on a large model according to an exemplary embodiment. DETAILED DESCRIPTION

[0032] The embodiments of the present application will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present application are shown in the accompanying drawings, it should be understood that the present application can be implemented in various forms and should not be construed as being limited to the embodiments described herein. Instead, these embodiments are provided to provide a more thorough and complete understanding of the present application. It should be understood that the drawings and embodiments of the present application are only for exemplary purposes and are not intended to limit the scope of protection of the present application.

[0033] It should be understood that the various steps described in the method implementation of the present application can be performed in different orders and / or performed in parallel. In addition, the method implementation may include additional steps and / or omit the steps shown. The scope of the present application is not limited in this respect.

[0034] The term "including" and its variations used herein are open inclusions, i.e., "including but not limited to". The term "based on" means "based at least in part on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". The relevant definitions of other terms will be given in the following description.

[0035] It should be noted that the concepts such as "first" and "second" mentioned in this application are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.

[0036] It should be noted that the modifications of "one" and "plurality" mentioned in the present application are illustrative rather than restrictive, and those skilled in the art should understand that unless otherwise clearly indicated in the context, it should be understood as "one or more".

[0037] The names of the messages or information exchanged between multiple devices in the embodiments of the present application are only used for illustrative purposes and are not used to limit the scope of these messages or information.

[0038] The specific implementation of the present application is described in detail below with reference to the accompanying drawings.

[0039] Figure 1 FIG. 1 is a block diagram of a customer resource management system based on a large model according to an exemplary embodiment. Figure 1 As shown, the system 100 includes:

[0040] The customer resource information acquisition module 101 is used to acquire a text description of a new customer resource;

[0041] A salesperson historical sales data acquisition module 102 is used to acquire historical sales data of candidate salespersons;

[0042] A semantic coding module 103, used for semantically coding the text description of the new customer resource and each sales data in the historical sales data of the candidate salesperson, respectively, to obtain a set of new customer resource semantic coding feature vectors and sales data semantic coding feature vectors;

[0043] A semantic association strengthening module 104, configured to perform semantic association strengthening on the set of sales data semantic coding feature vectors to obtain a set of strengthened sales data semantic coding feature vectors;

[0044] The resource push module 105 is used to determine whether to push the new customer resource to the candidate salesperson based on the fine-grained association matching between the new customer resource semantic encoding feature vector and the set of enhanced sales data semantic encoding feature vectors.

[0045] In response to the above technical problems, the technical concept of this application is to use natural language processing technology based on deep learning to perform semantic analysis on the sales staff's historical sales data and new customer resources, to mine the sales staff's professional experience information and the potential demand information of customer resources, and to conduct fine-grained correlation matching analysis on the two, so as to intelligently judge whether the new customer resources match the sales staff's capabilities and style, thereby achieving accurate push and optimized management of customer resources. In this way, the efficiency and accuracy of customer resource allocation can be significantly improved, the cost of ineffective communication in the sales process can be reduced, customer satisfaction can be improved, and enterprises can achieve sustained growth in sales performance.

[0046] Based on this, in the technical solution of the present application, first, a text description of new customer resources and historical sales data of candidate sales personnel are obtained. Among them, the text description of the new customer resources may include information such as the customer's basic information, business needs, industry, demand description, and interest preferences, while the historical sales data of the candidate sales personnel covers the sales personnel's past success stories, sales strategies, customer feedback and other information, reflecting the sales personnel's ability level, professional knowledge, and the types of customers they are good at handling. When allocating new customer resources, by considering the sales personnel's historical sales data, allowing the most professional and experienced sales personnel to adapt to the potential demand patterns of customer resources, the successful conversion rate of new customer resources can be effectively improved.

[0047] Next, in order to achieve a deep understanding of the text description of the new customer resources and the historical sales data of the candidate sales personnel at the semantic level, the present application further uses a pre-trained large-scale language model to semantically encode the text description of the new customer resources and each sales data in the historical sales data of the candidate sales personnel, so as to fully capture the deep semantic information such as themes, emotions and contextual semantic associations in the text data, and convert the unstructured text information into a structured vector representation, unify the data representation form, and obtain a set of semantic encoding feature vectors of new customer resources and semantic encoding feature vectors of sales data. In an embodiment of the present application, pre-trained language models such as BERT and GPT can be used to implement this semantic encoding process.

[0048] In one embodiment of the present application, the semantic association enhancement module is used to: perform a semantic association enhancement module based on feature energy level measurement association radiation on the set of semantic encoding feature vectors of the sales data to obtain the set of enhanced semantic encoding feature vectors of the sales data.

[0049] Then, it is considered that each sales data in the salesperson's historical sales data usually has a certain correlation. Therefore, in order to make full use of this correlation to achieve a global understanding of the salesperson's ability level and the type of expertise, the present application further adopts a semantic association enhancement module based on feature energy level measurement association radiation to perform semantic association enhancement encoding on the set of the sales data semantic encoding feature vectors. Specifically, the semantic association enhancement module quantifies the feature information density of each sales data and its feature radiation influence on other sales data by calculating the feature energy level coefficient of each sales data semantic encoding feature vector, and captures the local semantic dependency association relationship between each sales data by performing a one-dimensional convolution operation on each feature energy level coefficient, so as to generate attention weights, and enhance the distinguishability of the features by weighting and residual processing the set of the sales data semantic encoding feature vectors, and obtains a set of enhanced sales data semantic encoding feature vectors. In this way, it helps to extract the core capability characteristics and typical service models of sales personnel, and further improve the matching accuracy.

[0050] In one embodiment of the present application, the semantic association enhancement module includes: a characteristic energy level coefficient calculation unit, which is used to calculate the characteristic energy level coefficients of each sales data semantic coding feature vector in the set of sales data semantic coding feature vectors to obtain a set of characteristic energy level coefficients; a one-dimensional arrangement unit, which is used to arrange the set of characteristic energy level coefficients into a characteristic energy level coefficient input vector; a characteristic energy level radiation association coding unit, which is used to perform one-dimensional convolution encoding on the characteristic energy level coefficient input vector to obtain a characteristic energy level radiation association vector; a probabilistic mapping unit, which is used to perform point convolution encoding and probabilistic mapping based on the Sigmoid function on the characteristic energy level radiation association vector to obtain a characteristic energy level radiation association weight vector; a semantic enhancement unit, which is used to perform weighted processing and residual processing on the set of sales data semantic coding feature vectors using the characteristic energy level radiation association weight vector as a weight vector to obtain the set of enhanced sales data semantic coding feature vectors.

[0051] Furthermore, in one embodiment of the present application, the feature energy level coefficient calculation unit is used to: calculate the weighted sum between the variance of the sales data semantic encoding feature vector and the bias term to obtain a first intermediate term, and calculate the square of the difference between the maximum value of the sales data semantic encoding feature vector and its mean, the variance of the sales data semantic encoding feature vector and the bias term to obtain a second intermediate term; calculate the first intermediate term divided by the second intermediate term to obtain the feature energy level coefficient.

[0052] Specifically, the set of sales data semantic encoding feature vectors is processed by the following semantic association enhancement formula to obtain the set of enhanced sales data semantic encoding feature vectors, wherein the semantic association enhancement formula is:

[0053] X={x1,x2,...,x i ,...,x n}

[0054]

[0055] V e =(e1; e2; ...; e i ;...;e n )

[0056] V f =Conv 1×l (V e )

[0057] V c =Sigmoid(Conv 1×1 (V f ))

[0058] X′=X⊙V c +X

[0059] Wherein, X represents the set of the sales data semantic coding feature vectors, the value of n is the number of feature vectors in the set of the sales data semantic coding feature vectors, x1, x2, x i and x n represent the first, second, i-th and n-th sales data semantic encoding feature vectors in the set of sales data semantic encoding feature vectors, respectively, μ i is the feature mean of the semantic encoding feature vector of the i-th sales data, σ 2 i is the feature variance of the semantic encoding feature vector of the i-th sales data, max(x i ) represents the maximum value of each eigenvalue of the semantic encoding feature vector of the i-th sales data, b is the bias term, e1, e2, e i and e n represent the characteristic energy level coefficients of the first, second, i-th and n-th sales data semantic encoding feature vectors in the set of sales data semantic encoding feature vectors, V e Represents the characteristic energy level coefficient input vector, Conv 1×l (·) represents a one-dimensional convolution operation, l is the scale of the one-dimensional convolution kernel, V f represents the characteristic energy level radiation correlation vector, Conv 1×1 represents the point convolution operation, Sigmoid represents the Sigmoid function, V c represents the characteristic energy level radiation associated weight vector, ⊙ represents the dot product, and X′ represents the set of semantic encoding feature vectors of the enhanced sales data.

[0060] In one embodiment of the present application, the resource push module includes: a fine-grained association matching analysis unit, which is used to perform fine-grained association matching analysis on the set of the new customer resource semantic encoding feature vector and the enhanced sales data semantic encoding feature vector to obtain a query response adaptive matching result representation matrix; the query response adaptive matching result representation matrix is ​​input into a classifier-based customer resource management result generator to obtain a management result, and the management result is used to indicate whether to push new customer resources to alternative sales personnel.

[0061] Secondly, a matching analysis is further performed on the set of the new customer resource semantic encoding feature vectors and the enhanced sales data semantic encoding feature vectors to determine the degree of adaptation between the new customer resources and the alternative sales personnel. In the technical solution of the present application, in order to improve the matching accuracy, a feature query response adaptive encoder based on fine-grained decoupling is introduced to process both. Specifically, the feature query response adaptive encoder first performs fine-grained decoupling on each enhanced sales data semantic encoding feature vector, splits it into a sequence of local feature vectors, and uses a Bi-LSTM model to perform local information semantic association encoding on it, capturing the contextual information in each enhanced sales data semantic encoding feature vector to enhance its local information expression capability, so that each enhanced sales data semantic encoding feature vector has stronger expression capability and association at the local level. Next, the correlation between the new customer resource semantic encoding feature vector and the enhanced sales data semantic encoding feature vector after each local semantic association encoding is calculated, as well as the trace of the covariance matrix, to quantify the correlation between the two, and use this to generate attention weights. By weighting and calculating each covariance matrix, a query response adaptive matching result representation matrix is ​​obtained, which is used to comprehensively represent the overall matching degree between new customer resources and alternative sales personnel, and provide a decision-making basis for the precise allocation of new customer resources.

[0062] In one embodiment of the present application, the fine-grained association matching analysis unit is used to: use the new customer resource semantic encoding feature vector as a query feature vector, and input the query feature vector and the set of the enhanced sales data semantic encoding feature vector into a feature query response adaptive encoder based on fine-grained decoupling to obtain the query response adaptive matching result representation matrix.

[0063] Furthermore, in one embodiment of the present application, the fine-grained association matching analysis unit includes: a local semantic association strengthening subunit, which is used to perform local semantic association strengthening on each enhanced sales data semantic encoding feature vector in the set of enhanced sales data semantic encoding feature vectors to obtain a set of sales data fine-grained decoupled local semantic association feature vectors; and a covariance calculation subunit, which is used to calculate the covariance matrix between the query feature vector and each sales data fine-grained decoupled local semantic association feature vector in the set of sales data fine-grained decoupled local semantic association feature vectors to obtain a query response covariance matrix. A set of variance matrices; a matrix trace calculation subunit, used to calculate the matrix traces of each query response covariance matrix in the set of query response covariance matrices to obtain a set of query response trace measurement factors; a weighting subunit, used to input the set of query response trace measurement factors into a Sigmoid activation function to obtain a set of query response trace attention weights; an attention application subunit, used to calculate the weighted sum of the set of query response covariance matrices using each query response trace attention weight in the set of query response trace attention weights as a weight to obtain the query response adaptive matching result representation matrix.

[0064] Furthermore, in one embodiment of the present application, the local semantic association reinforcement subunit is used to: perform fine-grained decoupling of the enhanced sales data semantic encoding feature vector to obtain a sequence of enhanced sales data local semantic encoding feature vectors; input the sequence of enhanced sales data local semantic encoding feature vectors into a sequence semantic reinforcement encoder based on a Bi-LSTM model to obtain the sales data fine-grained decoupled local semantic association feature vector.

[0065] Specifically, the query feature vector and the set of the enhanced sales data semantic encoding feature vector are processed by the following query response encoding formula to obtain the query response adaptive matching result representation matrix:

[0066] S={S1,S2,...,S n}

[0067] S′ i =Decompose(S i )

[0068] {E i}=Bi_LSTM({S′ i})

[0069] {C i}={Cov(Q,E i )}

[0070] {τ i}={Tr(Ci )}

[0071] {α i}={σ(τ i )}

[0072]

[0073] Where S represents the set of semantic encoding feature vectors of the enhanced sales data, S1, S2, S i and S n represents the first, second, i-th and n-th enhanced sales data semantic coding feature vectors in the set of enhanced sales data semantic coding feature vectors respectively, the value of n is the number of enhanced sales data semantic coding feature vectors in the set of enhanced sales data semantic coding feature vectors, Decompose(·) represents decoupling processing, S′ i represents the sequence of local semantic encoding feature vectors of the i-th enhanced sales data, Bi_LSTM is the sequence semantic enhancement encoder based on the Bi-LSTM model, and E i represents the fine-grained decoupled local semantic association feature vector of the i-th sales data, Q represents the query feature vector, Cov(·) represents the covariance calculation function, C i represents the i-th query response covariance matrix, Tr(·) represents the matrix trace calculation function, τ i represents the metric factor of the i-th query response trace, σ represents the Sigmoid activation function, α i represents the attention weight of the ith query response trace, ⊙ represents the dot product, and m represents the query response adaptive matching result representation matrix.

[0074] Furthermore, the query response adaptive matching result representation matrix is ​​input into the classifier-based customer resource management result generator to obtain a management result, which is used to indicate whether to push new customer resources to candidate sales personnel. That is, a classifier model is used to construct a customer resource management result generator, and the query response adaptive matching result representation matrix is ​​subjected to binary classification through a fully connected network and a softmax function, so as to intelligently decide whether to push new customer resources to candidate sales personnel. In this way, it can be ensured that only customer resources that are highly matched with the salesperson's ability will be pushed, thereby avoiding invalid sales attempts, saving sales time, and improving sales efficiency.

[0075] In the technical solution of the present application, the new customer resource semantic encoding feature vector represents the contextual semantic features of the new customer resource text description, and each enhanced sales data semantic encoding feature vector in the set of enhanced sales data semantic encoding feature vectors respectively represents the contextual semantic enhanced association features of the text description of each sales data. In this way, when the new customer resource semantic encoding feature vector is used as a query feature vector, and the query feature vector and the set of enhanced sales data semantic encoding feature vectors are input into a feature query response adaptive encoder based on fine-grained decoupling, the obtained query response adaptive matching result representation matrix will cause insufficient query response coverage due to the scale mismatch between the query feature vector and the sales data feature vector fine-grained decoupling local association feature vector, thereby causing outlier class inference mapping deviation, affecting the accuracy of the management results obtained by the query response adaptive matching result representation matrix through the classifier-based customer resource management result generator.

[0076] Based on this, in a preferred example, inputting the query response adaptive matching result representation matrix into a classifier-based customer resource management result generator to obtain a management result includes:

[0077] Calculating an eigenvalue mean of the query response adaptive matching result representation matrix, and dividing the eigenvalue mean by a difference between a maximum eigenvalue and a minimum eigenvalue of the query response adaptive matching result representation matrix to obtain a query response adaptive matching result representation distribution representation value;

[0078] Subtracting the query response adaptive matching result representation distribution characterization value from one and dividing the result by the query response adaptive matching result representation distribution characterization value to obtain a query response adaptive matching result representation distribution modulation value;

[0079] Activating the query response adaptive matching result representation matrix through a probabilistic function to obtain a probabilistic query response adaptive matching result representation matrix;

[0080] After performing point subtraction between the probabilistic query response adaptive matching result representation matrix and the query response adaptive matching result representation distribution modulation value, taking the absolute value and calculating the negative of the logarithmic value with base 2 to obtain a probabilistic query response adaptive matching result representation distribution modulation information vector;

[0081] After dividing the query response adaptive matching result representation distribution characterization value by one minus the difference of each eigenvalue of the probabilistic query response adaptive matching result representation matrix, summing all eigenvalues ​​of the probabilistic query response adaptive matching result representation matrix and dividing the sum by the scale of the query response adaptive matching result representation matrix to obtain a probabilistic query response adaptive matching result representation distribution modulation bias value;

[0082] Performing a dot operation on the probabilistic query response adaptive matching result representation distribution modulation information vector and the product of the probabilistic query response adaptive matching result representation distribution modulation bias value and a weight as a hyperparameter to obtain an optimized query response adaptive matching result representation matrix; and

[0083] The optimized query response adaptive matching result representation matrix is ​​passed through the classifier-based customer resource management result generator to obtain the management result.

[0084] The optimized query response adaptive matching result representation matrix is ​​represented as:

[0085]

[0086] And among them:

[0087]

[0088] in, represents the characteristic mean of the query response adaptive matching result representation matrix, m max and m min represents the maximum eigenvalue and the minimum eigenvalue in the query response adaptive matching result representation matrix, respectively; p represents the query response adaptive matching result representation distribution representation value; M represents the probabilistic query response adaptive matching result representation matrix obtained after the query response adaptive matching result representation matrix is ​​activated by a probabilistic function; m i represents the i-th eigenvalue of the probabilistic query response adaptive matching result representation matrix, log represents a logarithmic function with base 2, ε is a weight as a hyperparameter, S is the scale of the query response adaptive matching result representation matrix, that is, the product of the width and height of the query response adaptive matching result representation matrix, M′ represents the optimized query response adaptive matching result representation matrix, represents positional subtraction, Represents positional addition.

[0089] That is, in the above preferred example, the probability information distribution planning based on the eigenvalues ​​of the query response adaptive matching result representation matrix is ​​performed through the Bernoulli probability modulation distribution of the query response adaptive matching result representation matrix relative to the eigenvalue distribution, and the probability reverse mapping of the probability characteristics of the query response adaptive matching result representation matrix as a whole is used as an extended coverage of the set mapping space of the query response adaptive matching result representation matrix to independently understand the interaction path between the intuitive probability information distribution and the abstract probability space mapping of the query response adaptive matching result representation matrix, so as to improve the accuracy of the management results obtained by the customer resource management result generator based on the classifier input by avoiding the counterfactual reasoning mapping of the outlier feature distribution of the query response adaptive matching result representation matrix to the class probability.

[0090] In summary, the above solution uses natural language processing technology based on deep learning to perform semantic analysis on the sales staff's historical sales data and new customer resources, dig out the sales staff's professional experience information and the potential demand information of customer resources, and conduct fine-grained correlation matching analysis on the two to intelligently determine whether the new customer resources match the sales staff's capabilities and style, thereby achieving accurate push and optimized management of customer resources. In this way, the efficiency and accuracy of customer resource allocation can be significantly improved, the cost of ineffective communication in the sales process can be reduced, customer satisfaction can be improved, and enterprises can achieve sustained growth in sales performance.

[0091] Figure 2 is a flowchart of a customer resource management method based on a large model according to an exemplary embodiment. Figure 2 As shown, the method includes:

[0092] Step S201: Obtain a text description of a new customer resource.

[0093] Step S202: Obtain historical sales data of candidate sales personnel;

[0094] Step S203: semantically encode the text description of the new customer resource and each sales data in the historical sales data of the candidate salesperson to obtain a set of new customer resource semantic encoding feature vectors and sales data semantic encoding feature vectors;

[0095] Step S204: performing semantic association enhancement on the set of sales data semantic coding feature vectors to obtain a set of enhanced sales data semantic coding feature vectors;

[0096] Step S205: Determine whether to push the new customer resource to the candidate salesperson based on the fine-grained association matching between the new customer resource semantic encoding feature vector and the set of enhanced sales data semantic encoding feature vectors.

[0097] In one embodiment of the present application, semantic association enhancement is performed on the set of sales data semantic coding feature vectors to obtain a set of enhanced sales data semantic coding feature vectors, including: performing a semantic association enhancement module based on feature energy level metric association radiation on the set of sales data semantic coding feature vectors to obtain the set of enhanced sales data semantic coding feature vectors.

[0098] Reference below Figure 3 , which shows a schematic diagram of the structure of an electronic device 600 suitable for implementing the embodiment of the present application. The terminal device in the embodiment of the present application may include but is not limited to mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 3 The electronic device shown is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0099] like Figure 3 As shown, the electronic device 600 may include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 601, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 602 or a program loaded from a storage device 608 into a random access memory (RAM) 603. In the RAM 603, various programs and data required for the operation of the electronic device 600 are also stored. The processing device 601, the ROM 602, and the RAM 603 are connected to each other via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0100] Typically, the following devices may be connected to the I / O interface 605: an input device 606 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 607 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 608 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 609. The communication device 609 may allow the electronic device 600 to communicate with other devices wirelessly or by wire to exchange data. Although Figure 3 The electronic device 600 is shown with various devices, but it should be understood that it is not required to implement or possess all the devices shown. More or fewer devices may be implemented or possessed instead.

[0101] In particular, according to an embodiment of the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present application includes a computer program product, which includes a computer program carried on a non-transitory computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication device 609, or installed from the storage device 608, or installed from the ROM 602. When the computer program is executed by the processing device 601, the above-mentioned functions defined in the method of the embodiment of the present application are executed.

[0102] It should be noted that the computer-readable medium mentioned above in the present application may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, device or device. In the present application, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program code. This propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The computer readable signal medium may also be any computer readable medium other than a computer readable storage medium, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer readable medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0103] In some embodiments, the client and the server may communicate using any currently known or future developed network protocol such as HTTP (HyperText Transfer Protocol), and may be interconnected with any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network ("LAN"), a wide area network ("WAN"), an internet (e.g., the Internet), and a peer-to-peer network (e.g., an ad hoc peer-to-peer network), as well as any currently known or future developed network.

[0104] The computer-readable medium may be included in the electronic device, or may exist independently without being incorporated into the electronic device.

[0105] Computer program code for performing the operations of the present application may be written in one or more programming languages ​​or a combination thereof, including, but not limited to, object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0106] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present application. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0107] The modules involved in the embodiments described in this application can be implemented by software or hardware. The name of the module does not limit the module itself in some cases. For example, the test parameter acquisition module can also be described as a "module for acquiring device test parameters corresponding to the target device".

[0108] The functions described above herein may be performed at least in part by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), and the like.

[0109] In the context of the present application, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0110] Figure 4 FIG. 1 is an application scenario diagram of a customer resource management system based on a large model according to an exemplary embodiment. Figure 4 As shown, in this application scenario, first, a text description of a new customer resource is obtained (for example, Figure 4 C1 shown in the figure); obtain the historical sales data of the candidate salesperson (for example, Figure 4 Then, the text description and historical sales data of the acquired new customer resources are input into a server (e.g., Figure 4 In S) shown in , the server is able to process the text description of the new customer resource and the historical sales data based on the customer resource management algorithm of the large model to determine whether to push the new customer resource to the alternative salesperson.

[0111] The above description is only a preferred embodiment of the present application and an explanation of the technical principles used. Those skilled in the art should understand that the scope of disclosure involved in the present application is not limited to the technical solution formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above disclosed concept. For example, the above features are replaced with the technical features with similar functions disclosed in this application (but not limited to) by each other to form a technical solution.

[0112] In addition, although each operation is described in a specific order, this should not be construed as requiring these operations to be performed in the specific order shown or to be performed in a sequential order. Under certain circumstances, multitasking and parallel processing may be advantageous. Similarly, although some specific implementation details are included in the above discussion, these should not be construed as limiting the scope of the application. Some features described in the context of a separate embodiment can also be implemented in a single embodiment in combination. On the contrary, the various features described in the context of a single embodiment can also be implemented in multiple embodiments individually or in any suitable sub-combination mode.

[0113] Although the subject matter has been described in language specific to structural features and / or method logic actions, it should be understood that the subject matter defined is not necessarily limited to the specific features or actions described above. On the contrary, the specific features and actions described above are merely exemplary forms of implementation. With respect to the apparatus in the above-mentioned embodiments, the specific manner in which each module performs the operation has been described in detail in the embodiments related to the method, and will not be described in detail here.

Claims

1. A customer resource management system based on a large model, characterized in that: include: The customer resource information acquisition module is used to obtain the text description of new customer resources; A salesperson historical sales data acquisition module is used to acquire the historical sales data of candidate salespersons; A semantic coding module, used for semantically coding the text description of the new customer resource and each sales data in the historical sales data of the candidate salesperson, respectively, to obtain a set of new customer resource semantic coding feature vectors and sales data semantic coding feature vectors; A semantic association strengthening module, used for performing semantic association strengthening on the set of sales data semantic coding feature vectors to obtain a set of strengthened sales data semantic coding feature vectors; The resource push module is used to determine whether to push the new customer resources to the candidate salesperson based on the fine-grained association matching between the semantic coding feature vector of the new customer resources and the set of semantic coding feature vectors of the enhanced sales data.

2. The customer resource management system based on a large model according to claim 1, characterized in that: The semantic association strengthening module is used to: The set of the sales data semantic coding feature vectors is subjected to a semantic association enhancement module based on feature energy level measurement association radiation to obtain the set of enhanced sales data semantic coding feature vectors.

3. The customer resource management system based on a large model according to claim 2 is characterized in that: The semantic association strengthening module includes: A characteristic energy level coefficient calculation unit, used for calculating the characteristic energy level coefficient of each sales data semantic encoding feature vector in the set of sales data semantic encoding feature vectors to obtain a set of characteristic energy level coefficients; A one-dimensional arrangement unit, used for arranging the set of characteristic energy level coefficients into a characteristic energy level coefficient input vector; A characteristic energy level radiation correlation coding unit, used for performing one-dimensional convolution coding on the characteristic energy level coefficient input vector to obtain a characteristic energy level radiation correlation vector; A probabilistic mapping unit, used for performing point convolution encoding and probabilistic mapping based on a Sigmoid function on the characteristic energy level radiation association vector to obtain a characteristic energy level radiation association weight vector; The semantic enhancement unit is used to use the characteristic energy level radiation associated weight vector as the weight vector to perform weighted processing and residual processing on the set of sales data semantic coding feature vectors to obtain the set of enhanced sales data semantic coding feature vectors.

4. The customer resource management system based on a large model according to claim 3 is characterized in that: The characteristic energy level coefficient calculation unit is used to: Calculate the weighted sum of the variance and the bias term of the sales data semantic encoding feature vector to obtain a first intermediate term, and calculate the square of the difference between the maximum value and the mean of the sales data semantic encoding feature vector, the variance of the sales data semantic encoding feature vector, and the bias term to obtain a second intermediate term; The first intermediate term divided by the second intermediate term is calculated to obtain the characteristic energy level coefficient.

5. The customer resource management system based on a large model according to claim 4 is characterized in that: The resource push module includes: A fine-grained association matching analysis unit, used for performing a fine-grained association matching analysis on the set of the new customer resource semantic encoding feature vector and the enhanced sales data semantic encoding feature vector to obtain a query response adaptive matching result representation matrix; The query response adaptive matching result representation matrix is ​​input into a classifier-based customer resource management result generator to obtain a management result, and the management result is used to indicate whether to push new customer resources to candidate sales personnel.

6. The customer resource management system based on a large model according to claim 5, characterized in that: The fine-grained association matching analysis unit is used to: The new customer resource semantic encoding feature vector is used as a query feature vector, and a set of the query feature vector and the enhanced sales data semantic encoding feature vector is input into a feature query response adaptive encoder based on fine-grained decoupling to obtain the query response adaptive matching result representation matrix.

7. The customer resource management system based on a large model according to claim 6 is characterized in that: The fine-grained association matching analysis unit includes: A local semantic association strengthening subunit, used for respectively performing local semantic association strengthening on each enhanced sales data semantic encoding feature vector in the set of enhanced sales data semantic encoding feature vectors to obtain a set of sales data fine-grained decoupled local semantic association feature vectors; A covariance calculation subunit, used for calculating the covariance matrix between the query feature vector and each sales data fine-grained decoupled local semantic association feature vector in the set of sales data fine-grained decoupled local semantic association feature vectors to obtain a set of query response covariance matrices; a matrix trace calculation subunit, configured to calculate the matrix trace of each query response covariance matrix in the set of query response covariance matrices to obtain a set of query response trace measurement factors; A weighting subunit, configured to input the set of query response trace measurement factors into a Sigmoid activation function to obtain a set of query response trace attention weights; The attention applying subunit is used to calculate the weighted sum of the set of query response covariance matrices using each query response trace attention weight in the set of query response trace attention weights as a weight to obtain the query response adaptive matching result representation matrix.

8. The customer resource management system based on a large model according to claim 7, characterized in that: The local semantic association strengthening subunit is used to: Fine-grained decoupling of the enhanced sales data semantic encoding feature vector is performed to obtain a sequence of enhanced sales data local semantic encoding feature vectors; The sequence of the enhanced sales data local semantic encoding feature vectors is input into a sequence semantic enhancement encoder based on the Bi-LSTM model to obtain the fine-grained decoupled local semantic association feature vector of the sales data.

9. A customer resource management method based on a large model, characterized in that: include: Get the text description of the new customer resource, Obtain historical sales data of candidate sales personnel; Semantically encoding the text description of the new customer resource and each sales data in the historical sales data of the candidate salesperson to obtain a set of new customer resource semantic encoding feature vectors and sales data semantic encoding feature vectors; Performing semantic association enhancement on the set of sales data semantic encoding feature vectors to obtain a set of enhanced sales data semantic encoding feature vectors; Based on the fine-grained association matching between the new customer resource semantic encoding feature vector and the set of enhanced sales data semantic encoding feature vectors, it is determined whether to push the new customer resource to the candidate salesperson.

10. The customer resource management method based on a large model according to claim 9, characterized in that: Performing semantic association enhancement on the set of sales data semantic encoding feature vectors to obtain a set of enhanced sales data semantic encoding feature vectors, including: The set of the sales data semantic coding feature vectors is subjected to a semantic association enhancement module based on feature energy level measurement association radiation to obtain the set of enhanced sales data semantic coding feature vectors.

Images