Medicine purchase data processing method and device, computer device, and storage medium

Abstract

The application discloses a medicine purchase data processing method and device, computer equipment and a storage medium. The medicine purchase data processing method can comprise the following steps: generating a preset number of first medicine purchase samples based on historical medicine purchase data, receiving label information of the first medicine purchase samples, correspondingly obtaining a preset number of second medicine purchase samples based on the label information, training a classification model by using the second medicine purchase samples, taking the trained classification model as an abnormal medicine purchase behavior model, taking a to-be-identified medicine purchase sample as an input of the abnormal medicine purchase behavior model, and determining whether the to-be-identified medicine purchase sample is an abnormal medicine purchase sample according to an output of the abnormal medicine purchase behavior model. Based on abnormal behavior mining and analysis of the historical medicine purchase data, the classification model is trained by using the labeled medicine purchase samples, and the abnormal medicine purchase behavior model is used to identify whether the current medicine purchase sample is an abnormal medicine purchase sample, so that the application has the advantages of high identification accuracy, rapidness, high intelligent degree and the like.

Description

Technical Field

[0001] The present invention relates to the field of big data processing technology. More specifically, the present invention can provide a method and device for processing drug purchase data, a computer device, and a storage medium. Background Art

[0002] With the expansion of the medical insurance system, the number of retail pharmacies and medical service institutions in cities has also increased. Currently, medical insurance fraud methods are constantly emerging, especially drug reselling. In drug reselling, the identities of some drug purchasers are used to purchase discounted drugs, and then the drugs are resold at a markup. To identify whether drug purchases are abnormal, traditional methods often rely on manual verification or electronic document verification. However, these existing methods are not only easy to circumvent, but the verification process is also very inefficient. Therefore, how to more accurately and quickly identify abnormal drug purchases has become a pressing technical problem and a research focus for those skilled in the art. Summary of the Invention

[0003] In order to solve the problems of existing methods for identifying abnormal drug purchasing behavior being easily circumvented and inefficient, the present invention can provide a drug purchasing data processing method and device, computer equipment, and storage medium to achieve technical goals such as improving the efficiency of identifying abnormal drug purchasing behavior and avoiding circumvention of identification means.

[0004] To achieve the above technical objectives, the present invention provides a method for processing drug purchase data, which may include but is not limited to one or more of the following steps.

[0005] A preset number of first drug purchase samples are generated based on historical drug purchase data, where the historical drug purchase data is drug purchase data within a first set time period.

[0006] Receive label information of a first drug purchase sample to obtain a preset number of second drug purchase samples based on the label information.

[0007] The second drug purchasing sample is used to train a classification model, and the trained classification model is used as an abnormal drug purchasing behavior model.

[0008] The drug purchase sample to be identified is used as the input of the abnormal drug purchase behavior model, and whether the drug purchase sample to be identified is an abnormal drug purchase sample is determined according to the output of the abnormal drug purchase behavior model.

[0009] In order to achieve the above technical objectives, the present invention can also provide a drug purchase data processing device, which specifically includes but is not limited to a drug purchase sample generation module, a drug purchase sample processing module, a classification model training module and an abnormal behavior recognition module.

[0010] The drug purchase sample generation module is used to generate a preset number of first drug purchase samples based on historical drug purchase data, where the historical drug purchase data is drug purchase data within a first set time period.

[0011] The medicine purchasing sample processing module is used to receive the label information of the first medicine purchasing sample and obtain a preset number of second medicine purchasing samples based on the label information.

[0012] The classification model training module is used to train the classification model using the second drug purchasing sample, and to use the trained classification model as an abnormal drug purchasing behavior model.

[0013] The abnormal behavior identification module is used to use the drug purchase sample to be identified as the input of the abnormal drug purchase behavior model, and to determine whether the drug purchase sample to be identified is an abnormal drug purchase sample based on the output of the abnormal drug purchase behavior model.

[0014] In order to achieve the above-mentioned technical objectives, an embodiment of the present invention can also provide a computer device, which includes a memory and a processor. The memory stores computer-readable instructions. When the computer-readable instructions are executed by the processor, the processor can execute the steps of the drug purchase data processing method described in any embodiment of the present invention.

[0015] In order to achieve the above-mentioned technical objectives, an embodiment of the present invention can also provide a storage medium storing computer-readable instructions. When the computer-readable instructions are executed by one or more processors, the one or more processors execute the steps of the drug purchase data processing method described in any embodiment of the present invention.

[0016] To achieve the above technical objectives, the present invention can also provide a computer program product. When the instructions in the computer program product are executed by a processor, the steps of the drug purchase data processing method described in any embodiment of the present invention are executed.

[0017] The beneficial effects of the present invention are:

[0018] Based on the mining and analysis of abnormal behavior in historical drug purchase data, the present invention specifically trains a classification model using labeled drug purchase samples. The trained abnormal drug purchase behavior model then automatically identifies whether the current drug purchase sample is abnormal. This demonstrates the present invention's ability to accurately, quickly, and intelligently identify abnormal drug purchase behavior, achieving a highly efficient and difficult-to-circumvent identification. This invention effectively combines big data and artificial intelligence technologies, accurately identifying drug reselling through big data mining and analysis, and promptly notifying relevant personnel to rapidly address the situation. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 A flow chart of a method for processing drug purchase data in one or more embodiments of the present invention is shown.

[0020] Figure 2 A flow chart of a method for processing drug purchase data based on a single classification model in one or more embodiments of the present invention is shown.

[0021] Figure 3 A flow chart of a method for processing drug purchase data for abnormality determination in one or more embodiments of the present invention is shown.

[0022] Figure 4 A flow chart of a method for processing drug purchase data based on structured drug purchase data in one or more embodiments of the present invention is shown.

[0023] Figure 5 A flow chart of a method for processing drug purchase data based on a drug purchase behavior dataset in one or more embodiments of the present invention is shown.

[0024] Figure 6 A flow chart of a method for processing drug purchase data by filtering and uniformly representing a drug purchase behavior dataset in one or more embodiments of the present invention is shown.

[0025] Figure 7 A schematic structural diagram of a drug purchase data processing device in one or more embodiments of the present invention is shown.

[0026] Figure 8 A schematic diagram showing the internal structure of a computer device in one or more embodiments of the present invention is shown. DETAILED DESCRIPTION

[0027] The following is a detailed explanation and description of a drug purchase data processing method and device, computer equipment, and storage medium provided by an embodiment of the present invention in conjunction with the drawings in the specification.

[0028] like Figure 1 shown, and can be combined Figures 2 to 6 One or more embodiments of the present invention may provide a method for processing drug purchase data. The data processing method includes but is not limited to one or more of the following steps, which are described in detail below.

[0029] Step 100: Generate a preset number of first drug purchase samples based on the historical drug purchase data. In embodiments of the present invention, the historical drug purchase data refers to drug purchase data within a first set time period. This historical drug purchase data can be provided by all retail pharmacies in a city that have designated medical insurance services. Data can be automatically collected by connecting to various drug purchase management platforms, enabling the collection of all drug purchase behavior data from a designated list of pharmacies. The first set time period in embodiments of the present invention can be, for example, the past year or six months, but is not limited thereto.

[0030] like Figure 4 As shown, in an embodiment of the present invention, generating a preset number of first drug purchase samples based on historical drug purchase data includes but is not limited to the following steps 101 and 102.

[0031] Step 101: vectorize the historical drug purchase data to obtain structured drug purchase data in vector form. The embodiment of the present invention structures the collected historical drug purchase data, specifically in vector form, for example, p∈R d The present invention can obtain a series of structured drug purchasing data through this method, that is, a series of drug purchasing behavior samples: p1, p2, ..., p N , N can represent the total number of samples.

[0032] Wherein, p represents structured data in vector form, R represents a specific vector, and d represents the data dimension. One or more embodiments of the present invention may include performing vector processing on historical drug purchase data: vectorizing the historical drug purchase data according to the dimensions of drug purchaser information, drug store information, and drug purchase item information, where d = d1 + d2 + d3, where d1 can be used to represent the drug purchaser information dimension, d2 can be used to represent the drug store information dimension, and d3 can be used to represent the drug purchase item information dimension. Specifically, in the embodiment of the present invention, p = [a, b, c], Among them, a represents the information vector of the drug purchaser, b represents the information vector of the pharmacy, and c represents the information vector of the drug purchase details.

[0033] Specifically, the drug purchaser information in this embodiment may include but is not limited to name, age, social security card information, unit information, address information, etc. The pharmacy information in this embodiment may include but is not limited to the pharmacy's address information, business management license information, etc. The drug purchase details information in this embodiment may include but is not limited to the date and time of purchase, the name and quantity of the purchased drugs, etc. Of course, the present invention is not limited to the above examples in specific application scenarios, and may also include other drug purchase related information.

[0034] The following example uses the drug purchase details information vector as an example: In this embodiment, the names of the drugs in the drug purchase details can be coded according to the specified drug list, and the number of drugs purchased can be marked at the corresponding coded position. For example, if there are ten drugs under supervision, c = [t, c i ], c i =[0,0,0,0,0,0,0,0,0,0], where t represents the date and time of drug purchase, c i Indicates the type and quantity of the purchased medicines; if two quantities of the first medicine and one quantity of the second medicine are purchased in a certain purchase, then c is used to indicate the type and quantity of the purchased medicines. i =[2,1,0,0,0,0,0,0,0,0]; if the time is 12:56 on December 15, 2021, then t=[20211215,1256]. Of course, the present invention is not limited to the above example.

[0035] Step 102 : Generate a preset number of first drug purchase samples based on the structured drug purchase data. That is, the embodiment of the present invention determines the first drug purchase samples based on the obtained structured drug purchase data.

[0036] like Figure 5 As shown, one or more embodiments of the present invention may generate a preset number of first drug purchase samples based on structured drug purchase data, which may include but is not limited to the following steps 103 and 104, which are specifically described as follows.

[0037] Step 103: Obtain the social security card information associated with the historical drug purchase data; divide the structured drug purchase data by category based on the social security card information to obtain multiple drug purchase behavior data sets. Different drug purchase behavior data sets correspond to different social security card information. In the present invention, "social security card information" can be replaced with "medical insurance card information." Both social security card information and medical insurance card information can be used to represent medical insurance-related information. In this embodiment, they can be used to divide the structured drug purchase data by category. Combined with the above-mentioned vector representation, this embodiment of the present invention can divide the structured drug purchase data by the social security card information contained in vector a.

[0038] For structured drug purchase data p1, p2,…, p N In the embodiment of the present invention, the social security card information corresponding to the drug purchasing behavior is archived, that is, the associated individual is archived to achieve the purpose of dividing the structured drug purchasing data according to categories. The multiple drug purchasing behavior data sets obtained after archiving are f1, f2, ..., f M , where M is used to represent the number of drug purchase behavior datasets. For any drug purchase behavior dataset f i , including at least one structured drug purchasing data, for example, f2 = {p3, p5}, of course not limited to this.

[0039] Step 104 : Generate a preset number of first drug purchasing samples based on the drug purchasing behavior dataset. That is, the embodiment of the present invention determines the first drug purchasing samples based on the obtained drug purchasing behavior dataset.

[0040] like Figure 6 As shown, one or more embodiments of the present invention may generate a preset number of first drug purchase samples based on the drug purchase behavior dataset, which may include but is not limited to the following steps 105 and 106, as described in detail below.

[0041] Step 105, screen the drug purchasing behavior dataset to filter out the drug purchasing behavior dataset within the second set time period from the drug purchasing behavior dataset. In this embodiment, the second set time period is less than the first set time period, and the second set time period can be, for example, one month. In the abnormal drug purchasing behavior that has occurred, in order to seek huge profits from drug resale, the abnormal drug purchasing behavior is often manifested as behavior within a short period of time, and a certain drug is purchased multiple times in different pharmacies with the same social security card or even different identity social security cards. The present invention normalizes the drug purchasing behavior dataset by setting certain time rules to ensure that the drug purchasing behavior dataset has the same time length. It can be seen that the present invention can select the drug purchasing behavior dataset within a short period of time by screening the drug purchasing behavior dataset, thereby more accurately describing the drug purchasing behavior in a short period of time, making the drug purchasing behavior dataset more targeted, and thus making the abnormal drug purchasing behavior model trained subsequently have greater reliability and robustness.

[0042] Step 106: The drug purchase behavior dataset obtained by screening within the second set time period is uniformly represented in a preset format to obtain a preset number of first drug purchase samples. The embodiment of the present invention can obtain the drug purchase behavior dataset f in this way. i Unified expression of i .

[0043] Specifically, for the multiple drug purchase behavior data sets in the embodiment of the present invention, f1, f2, ..., f M , this embodiment follows the following preset format s i =[l i ,t i ,b i ] are uniformly represented to obtain the first drug purchase samples s1, s2,…, s M . Among them l i The pharmacy information expression represents the set of drug purchase behaviors coded according to the city pharmacies, t i Indicates the aggregate time information of drug purchasing behavior in the past month, b i Indicates the detailed information of medicine purchases in the past month.

[0044] Based on the above-mentioned improved technical solution, the present invention performs vectorization processing on the drug purchase big data, associates it with social security card information, filters the data according to set time rules, and uniformly represents it, thereby realizing the preprocessing of historical drug purchase big data, and then obtaining a preset number of first drug purchase samples. The present invention prepares a large number of high-quality training samples for the classification model in the above manner, providing original data support for the effective training of the classification model.

[0045] Step 200, receiving the label information of the first drug purchase sample, so as to obtain a preset number of second drug purchase samples based on the label information. The label information includes but is not limited to label information that characterizes sample abnormalities and label information that characterizes sample normality. Based on the received label information, this embodiment labels a preset number of first drug purchase samples to obtain a preset number of second drug purchase samples; based on the label information that characterizes sample abnormalities, this embodiment can label the first drug purchase samples with abnormalities, and based on the label information that characterizes sample normality, this embodiment can label the first drug purchase samples with normalities. For the first drug purchase samples s1, s2,…, s M , the present invention can obtain abnormal drug purchasing behaviors that have occurred, and drug purchasing behaviors other than abnormal drug purchasing behaviors are regarded as normal drug purchasing behaviors. Specifically, the present invention can represent abnormal labels through the label "1" and normal labels through the label "0"; for example, if the label of s2 is 0, it can be said that s2 indicates that the drug purchasing sample belongs to a normal consumer. The present invention can achieve the labeling of the first drug purchasing sample by manual labeling or automatic labeling using an existing labeling model. Of course, it is not limited to this, and the purpose of sample labeling of the present invention can be achieved.

[0046] Step 300: Use the second drug purchase sample to train the classification model, and use the trained classification model as an abnormal drug purchase behavior model. This embodiment of the present invention can use the abnormal drug purchase behavior model to distinguish the drug purchase behavior to be identified to determine whether it is abnormal drug purchase behavior.

[0047] like Figure 2 As shown, the classification model used in this embodiment is specifically a single classification model. One or more embodiments of the present invention use the second drug purchase sample to train the classification model, including but not limited to steps 301 and 302.

[0048] Step 301: Train the single-classification model using a second drug purchase sample obtained based on label information representing sample anomalies. During single-classification model training, this embodiment specifically aggregates features of abnormal drug purchase behavior through a single-classification approach. Using the second drug purchase sample obtained based on label information representing sample anomalies, the single-classification model learns drug purchase behavior over a short period of time. This invention effectively extracts reliable features for discriminating drug reselling behavior through the single-classification model, and identifies drug purchase behavior based on these reliable features.

[0049] Specifically, the embodiment of the present invention focuses on analyzing and mining the characteristics of drug purchase behavior with the label "1". For a single-classification model, the embodiment of the present invention can be described by the following objective function, which is described in detail below.

[0050]

[0051] Among them, W represents the model parameter matrix, M represents the number of second drug purchase samples, Φ(W,s i ) represents the matrix transformation function, s i represents the second drug purchase sample, C represents the mean sum of the feature values ​​of samples with label “1”, λ represents the hyperparameter, and ||W||1 represents the calculation of the norm of |W|.

[0052] Based on the above objective function, the present invention can learn samples with the label "1" by establishing a model W. More specifically, this embodiment establishes a model W using a forward deep neural network. The specific network expression is W = W1×W2×W3. It can be seen that the present invention can specifically use a three-layer forward neural network W1, W2, and W3 to learn the objective function, wherein the matrix transformation function Φ(W,s i ) can be specifically W1×W2×W3×s i , of course it is not limited to this.

[0053] In step 302, the trained single-classification model is used as the abnormal drug purchasing behavior model. It can be seen that the abnormal drug purchasing behavior model in this embodiment is a trained single-classification model.

[0054] Based on the above-mentioned improved technical solution, the present invention only trains a single-classification model through drug purchase samples with abnormal labels. Compared with the method of training a two-classification model through drug purchase samples with both normal and abnormal labels, the present invention realizes the feature aggregation of abnormal drug purchase behaviors through the above method, thereby effectively avoiding the influence of the behavior of a variety of normal drug buyers on the model training error. It can be seen that the improved solution of the present invention has the advantages of more targeted model training, less data volume and lower training cost. It should be understood that, based on the technical content disclosed in the present invention, other existing classification models can also be selected based on the solution provided by the present invention. However, this equivalent replacement or adjustment based on the substantive content of the present invention is also within the scope of protection of the present invention. Based on the substantive content disclosed in the present invention, the specific training process of the classification model can be selected from the existing classification model training scheme according to the needs of the actual application scenario of the present invention, such as the scale of the model or the sample size, so as to achieve the purpose of model training. The present invention will not elaborate on this.

[0055] In step 400, the drug purchasing sample to be identified is used as the input of the abnormal drug purchasing behavior model, and whether the drug purchasing sample to be identified is an abnormal drug purchasing sample is determined based on the output of the abnormal drug purchasing behavior model; the drug purchasing sample to be identified often includes multiple purchase records corresponding to a certain social security card. Specifically, the embodiment of the present invention utilizes the abnormal drug purchasing behavior model to output the abnormality degree to determine whether the current drug purchasing sample to be identified is an abnormal drug purchasing sample based on the abnormality degree. The higher the abnormality degree in this embodiment, the higher the possibility that the purchasing behavior in the drug purchasing sample to be identified involves drug reselling, and the lower the abnormality degree, the lower the possibility that the purchasing behavior in the drug purchasing sample to be identified involves drug reselling. Of course, the present invention can also express the possibility that the purchasing behavior involves drug reselling through normality. For example, the normality can be the inverse of the abnormality degree, so as to achieve the technical purpose of the present invention.

[0056] In specific implementation, when a drug buyer purchases a drug, the embodiment of the present invention first obtains the drug purchase behavior data within the historical time period (T) through the drug purchase management platform, such as the drug purchase behavior data of the shopper in the past month. The drug purchase sample to be identified corresponding to the current drug purchase behavior is obtained through the process of processing the drug purchase data by the embodiment of the present invention. In this embodiment, the drug purchase sample to be identified can be represented by s. Combined with the model Φ(W,s i ), the abnormality calculation formula of this embodiment is: 1 / ||Φ(W,s)-C|| 2 The higher the abnormality calculated by this formula, the higher the possibility that the current shopper is involved in drug reselling when purchasing the current drug.

[0057] like Figure 3As shown, the embodiment of the present invention determines whether the to-be-identified drug purchase sample is an abnormal drug purchase sample based on the output of the abnormal drug purchase behavior model, including but not limited to steps 401 to 403.

[0058] In step 401, the abnormality degree output by the abnormal drug purchasing behavior model is compared with a threshold value. Based on the comparison result, it is determined whether the current drug purchasing sample to be identified is an abnormal drug purchasing sample. The specific value of the threshold value can be set according to actual application conditions and is not limited in this embodiment.

[0059] In step 402, if the abnormality is greater than or equal to the threshold, the drug purchase sample to be identified may be determined to be an abnormal drug purchase sample.

[0060] Optionally, one or more embodiments of the present invention may further include, after determining that the to-be-identified drug purchase sample is an abnormal drug purchase sample, issuing at least one alert message for alerting of abnormal drug purchase behavior. The present invention issues an alert when the model output is abnormal, thereby notifying relevant personnel as soon as possible.

[0061] In step 403, if the abnormality is less than the threshold, the drug purchase sample to be identified can be determined to be a normal drug purchase sample.

[0062] like Figure 7 As shown, based on the same inventive technical concept as the drug purchase data processing method of the present invention, one or more embodiments of the present invention can also provide a drug purchase data processing device.

[0063] The drug purchase data processing device may include but is not limited to a drug purchase sample generation module 500, a drug purchase sample processing module 600, a classification model training module 700 and an abnormal behavior recognition module 800, which are described in detail as follows.

[0064] The drug purchase sample generation module 500 is configured to generate a preset number of first drug purchase samples based on historical drug purchase data, where the historical drug purchase data is drug purchase data within a first set time period.

[0065] Optionally, the drug purchase sample generation module 500 can be used to vectorize historical drug purchase data to obtain structured drug purchase data in vector form; the drug purchase sample generation module 500 is used to generate a preset number of first drug purchase samples based on the structured drug purchase data.

[0066] Specifically, in one or more embodiments of the present invention, the drug purchase sample generation module 500 is used to vectorize historical drug purchase data according to the drug purchaser information dimension, the drug store information dimension, and the drug purchase item information dimension, but is certainly not limited thereto.

[0067] Optionally, the drug purchase sample generation module 500 in this embodiment of the present invention is configured to obtain social security card information associated with historical drug purchase data and to classify the structured drug purchase data by category based on the social security card information to generate multiple drug purchase behavior datasets. Different drug purchase behavior datasets correspond to different social security card information. The drug purchase sample generation module 500 is configured to generate a preset number of first drug purchase samples based on the drug purchase behavior datasets.

[0068] Optionally, the drug purchasing sample generation module 500 is specifically configured to filter the drug purchasing behavior dataset to select drug purchasing behavior datasets within a second set time period, wherein the second set time period is shorter than the first set time period. The drug purchasing sample generation module 500 is configured to uniformly represent the drug purchasing behavior datasets within the second set time period in a preset format to obtain a preset number of first drug purchasing samples.

[0069] The medicine purchasing sample processing module 600 is configured to receive label information of a first medicine purchasing sample and obtain a preset number of second medicine purchasing samples based on the label information.

[0070] The classification model training module 700 is used to train the classification model using the second drug purchasing sample, and to use the trained classification model as an abnormal drug purchasing behavior model.

[0071] Optionally, the classification model in the embodiment of the present invention is a single classification model. The classification model training module 700 is specifically used to train the single classification model using the second drug purchase sample obtained based on the label information representing the sample abnormality, and to use the trained single classification model as the abnormal drug purchase behavior model.

[0072] The abnormal behavior identification module 800 is used to use the drug purchase sample to be identified as the input of the abnormal drug purchase behavior model, and to determine whether the drug purchase sample to be identified is an abnormal drug purchase sample based on the output of the abnormal drug purchase behavior model.

[0073] Optionally, the abnormal behavior identification module 800 in the embodiment of the present invention can be specifically used to compare the abnormality degree output by the abnormal drug purchasing behavior model with a threshold value. The abnormal behavior identification module 800 is used to determine that the drug purchasing sample to be identified is an abnormal drug purchasing sample based on the abnormality degree being greater than or equal to the threshold value; or, the abnormal behavior identification module 800 is used to determine that the drug purchasing sample to be identified is a normal drug purchasing sample based on the abnormality degree being less than the threshold value.

[0074] Optionally, the drug purchase data processing device in some embodiments of the present invention includes a drug purchase abnormality reminder module. The drug purchase abnormality reminder module is configured to issue at least one reminder message for abnormal drug purchase behavior based on the drug purchase sample to be identified as an abnormal drug purchase sample.

[0075] like Figure 8 As shown, the method for processing drug purchase data can be based on the same inventive technical concept. One or more embodiments of the present invention can also provide a computer device comprising a memory and a processor. The memory stores computer-readable instructions. When the computer-readable instructions are executed by the processor, the processor executes the steps of the method for processing drug purchase data in any embodiment of the present invention. It is understood that the detailed process of the method for processing drug purchase data has been fully described in this specification and will not be repeated here.

[0076] like Figure 8 As shown, the method for processing drug purchase data can be based on the same inventive technical concept. One or more embodiments of the present invention can also provide a storage medium storing computer-readable instructions. When executed by one or more processors, these computer-readable instructions cause the one or more processors to perform the steps of the method for processing drug purchase data in any embodiment of the present invention. It should be understood that the detailed process of the method for processing drug purchase data of the present invention has been fully described in this specification and will not be repeated here.

[0077] Based on the same technical concept as the drug purchase data processing method, one or more embodiments of the present invention may also provide a computer program product. When the instructions in the computer program product are executed by a processor, the steps of the drug purchase data processing method described in any embodiment of the present invention are performed. It should be understood that the detailed process of the drug purchase data processing method involved in the embodiments of the present invention has been fully described in this specification and will not be repeated here.

[0078] The technical solution provided by the present invention can be used in a pharmacy digital management platform (or called a "pharmacy information management system"). The abnormal drug purchasing behavior model can run on the application layer of the pharmacy digital management platform. The pharmacy digital management platform can be connected to all relevant drug purchasing management platforms in the city that have medical insurance designated services, thereby using the pharmacy digital management platform to intelligently and quickly judge abnormal drug purchasing behavior, and to discover the drug reselling behavior using medical insurance cards, so as to achieve the purpose of curbing drug reselling behavior.

[0079] The logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing the logical functions, and can be embodied in any computer-readable storage medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device). For purposes of this specification, a "computer-readable storage medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable storage media include the following: an electrical connection with one or more wires (electronic device), a portable computer disk cartridge (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM, or flash memory), a fiber optic device, and a portable compact disc read-only memory (CDROM). In addition, the computer-readable storage medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering, or processing in another suitable manner as necessary, and then stored in a computer memory.

[0080] It should be understood that various parts of the present invention can be implemented using hardware, software, firmware, or a combination thereof. In the above-described embodiments, multiple steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application-specific integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0081] In the description of this specification, the description with reference to the terms "this embodiment", "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art can combine and combine different embodiments or examples described in this specification and the features of different embodiments or examples, unless they are contradictory.

[0082] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of the technical features being referred to. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one such feature. In the description of the present invention, "plurality" means at least two, such as two, three, etc., unless otherwise specifically defined.

[0083] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent replacements and simple improvements made to the essential contents of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for processing drug purchase data, characterized in that: include: generating a preset number of first drug purchase samples based on historical drug purchase data, wherein the historical drug purchase data is the drug purchase data of all drugs under supervision collected from various pharmacies within a first set time period; receiving label information of a first drug purchase sample, performing abnormal labeling processing and normal labeling processing on the first drug purchase sample, so as to obtain a preset number of second drug purchase samples correspondingly based on the label information; The classification model is trained using the second drug purchase sample obtained based on the label information representing the sample abnormality to mine and analyze the abnormal behavior of the historical drug purchase data, and the trained classification model is used as the abnormal drug purchase behavior model. The objective function of the abnormal drug purchase behavior model is: Among them, W represents the model parameter matrix, M represents the number of second drug purchase samples, Φ(W,s i ) represents the matrix transformation function, s i represents the second drug purchase sample, C represents the sum of the mean characteristic values ​​of the abnormal labels, λ represents the hyperparameter, ||W||1 represents the calculation of the norm of |W|; W=W1×W2×W3, W1, W2, W3 represent the three-layer forward neural network, Φ(W,s i ) is W1×W2×W3×s i ; The drug purchasing sample to be identified is used as the input of the abnormal drug purchasing behavior model, and whether the drug purchasing sample to be identified is an abnormal drug purchasing sample is determined according to the output of the abnormal drug purchasing behavior model, so as to lock in the drug reselling behavior.

2. The drug purchase data processing method according to claim 1, characterized in that: The classification model is a single-classification model.

3. The method for processing drug purchase data according to claim 1, characterized in that: Determining whether the to-be-identified drug purchase sample is an abnormal drug purchase sample based on the output of the abnormal drug purchase behavior model includes: Comparing the abnormality degree output by the abnormal drug purchasing behavior model with a threshold; If the abnormality is greater than or equal to the threshold, determining that the drug purchase sample to be identified is an abnormal drug purchase sample; If the abnormality is less than the threshold, the drug purchase sample to be identified is determined to be a normal drug purchase sample.

4. The method for processing drug purchase data according to claim 1, wherein: Generating a preset number of first drug purchase samples based on historical drug purchase data includes: Performing vectorization processing on the historical drug purchase data to obtain structured drug purchase data in vector form; The preset number of first drug purchasing samples is generated based on the structured drug purchasing data.

5. The method for processing drug purchase data according to claim 4, characterized in that: Generating the preset number of first drug purchase samples based on the structured drug purchase data includes: Obtaining social security card information associated with the historical drug purchase data; Dividing the structured drug purchasing data by category according to the social security card information to obtain multiple drug purchasing behavior data sets; Among them, different drug purchase behavior datasets correspond to different social security card information; The preset number of first drug purchasing samples are generated based on the drug purchasing behavior dataset.

6. The method for processing drug purchase data according to claim 5, characterized in that: Generating the preset number of first drug purchasing samples based on the drug purchasing behavior dataset includes: screening the drug purchasing behavior dataset to filter out drug purchasing behavior datasets within a second set time period from the drug purchasing behavior dataset; Wherein, the second set time period is shorter than the first set time period; The drug purchasing behavior data set within the second set time period is uniformly represented in a preset format to obtain the preset number of first drug purchasing samples.

7. The method for processing drug purchase data according to claim 4, characterized in that: The vectorization processing of the historical drug purchase data includes: The historical drug purchase data is represented by vectors according to the drug purchaser information dimension, the drug store information dimension, and the drug purchase item information dimension.

8. A drug purchase data processing device, characterized in that: include: a drug purchase sample generation module, configured to generate a preset number of first drug purchase samples based on historical drug purchase data, wherein the historical drug purchase data is drug purchase data of all regulated drugs collected from various pharmacies within a first set time period; a medicine purchase sample processing module, configured to receive label information of a first medicine purchase sample, and perform abnormal labeling and normal labeling processing on the first medicine purchase sample, so as to obtain a preset number of second medicine purchase samples corresponding to the label information; The classification model training module is used to train the classification model using the second drug purchase sample obtained based on the label information representing the sample abnormality, so as to mine and analyze the abnormal behavior of the historical drug purchase data, and to use the trained classification model as an abnormal drug purchase behavior model, wherein the objective function of the abnormal drug purchase behavior model is: Among them, W represents the model parameter matrix, M represents the number of second drug purchase samples, Φ(W,s i ) represents the matrix transformation function, s i represents the second drug purchase sample, C represents the sum of the mean characteristic values ​​of the abnormal labels, λ represents the hyperparameter, ||W||1 represents the calculation of the norm of |W|; W=W1×W2×W3, W1, W2, W3 represent the three-layer forward neural network, Φ(W,s i ) is W1×W2×W3×s i ; The abnormal behavior identification module is used to use the drug purchasing sample to be identified as the input of the abnormal drug purchasing behavior model, and to determine whether the drug purchasing sample to be identified is an abnormal drug purchasing sample based on the output of the abnormal drug purchasing behavior model, so as to lock in the drug reselling behavior.

9. A computer device, characterized in that: The method comprises a memory and a processor, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the processor executes the steps of the drug purchase data processing method according to any one of claims 1 to 7.

10. A storage medium storing computer-readable instructions, characterized in that: When the computer-readable instructions are executed by one or more processors, the one or more processors execute the steps of the drug purchase data processing method according to any one of claims 1 to 7.

Images