CN111696675B

CN111696675B - User data classification method and device based on Internet of things data and computer equipment

Info

Publication number: CN111696675B
Application number: CN202010442300.0A
Authority: CN
Inventors: 黄德生
Original assignee: Shenzhen Saiante Technology Service Co Ltd
Current assignee: Shenzhen Saiante Technology Service Co Ltd
Priority date: 2020-05-22
Filing date: 2020-05-22
Publication date: 2023-09-19
Anticipated expiration: 2040-05-22
Also published as: WO2021232591A1; CN111696675A

Abstract

The invention discloses a user data classification method, a device, computer equipment and a storage medium based on internet of things data, which are applied to the intelligent medical field of intelligent cities and relate to data processing and blockchain. According to the method, the user medical data is quickly acquired based on the Internet of things equipment, the optimal treatment scheme data is quickly and accurately recommended based on the historical medical data, and the cost for acquiring the optimal treatment scheme data is reduced.

Description

User data classification method and device based on Internet of things data and computer equipment

Technical Field

The present invention relates to the field of data processing and blockchain technologies, and in particular, to a method and apparatus for classifying user data based on internet of things data, a computer device, and a storage medium.

Background

Currently, when a patient goes to a hospital to see a doctor (especially, chronic diseases such as hypertension, diabetes, cardiovascular and cerebrovascular diseases, etc.), the patient generally goes to a hospital site to measure the physical sign data, the patient manually reads and manually records the physical sign data and then informs a doctor, and the doctor records the physical sign data into a medical system server to be used as the historical medical data of the patient. However, the frequent going to hospital for manual measurement not only consumes time for the patient, but also occupies medical resources such as doctors, and the like, resulting in a longer acquisition period of medical data of chronic diseases.

And when a doctor recommends a subsequent treatment scheme according to physical sign data measured by a user in a hospital field, the doctor is generally based on the self experience of the doctor, and has higher requirements on the professional level of the doctor, namely, the labor cost for acquiring more accurate treatment scheme data is higher.

Disclosure of Invention

The embodiment of the invention provides a user data classification method, a device, computer equipment and a storage medium based on Internet of things data, and aims to solve the problems that in the prior art, acquisition of physical sign data of a patient user needs to go to a hospital site for manual measurement, and when a follow-up treatment scheme is recommended based on the measured physical sign data, the requirement on the professional level of a doctor is high, so that the acquisition period of medical data of chronic diseases is long, and the labor cost for acquiring accurate treatment scheme data is high.

In a first aspect, an embodiment of the present invention provides a method for classifying user data based on internet of things data, including:

judging whether current user medical data uploaded by the medical terminal of the Internet of things is received or not; the current user medical data comprise user identification data, drug injection data and physical sign measurement data;

if current user medical data uploaded by the medical terminal of the Internet of things is received, acquiring a historical medical data set corresponding to the identification data; wherein the historical medical data set comprises an electronic medical record data subset, an examination report data subset, a medication record data subset and a historical treatment scheme data subset;

invoking a preset first keyword set and a first extraction strategy corresponding to the first keyword set, and extracting a target data set in a historical medical data set corresponding to the identification data so as to be combined with the current user medical data to obtain current user input data;

the semantic vectors corresponding to the fields in the current user input data form a current user data input vector, the current user data input vector is input into a pre-trained convolutional neural network model, and a classification result corresponding to the current user data input vector is obtained;

Judging whether the treatment path data category corresponding to the classification result is greater than 1;

if the treatment path data types corresponding to the classification results are larger than 1, standard user data input vectors corresponding to the treatment path data types corresponding to the classification results are obtained; and

and calculating vector similarity between the standard user data input vector corresponding to the classification result and the current user data input vector, and acquiring the standard user data input vector corresponding to the maximum vector similarity and the corresponding treatment path data as optimal treatment path data.

In a second aspect, an embodiment of the present invention provides a user data classification device based on internet of things data, including:

the current data receiving unit is used for judging whether current user medical data uploaded by the medical terminal of the Internet of things are received or not; the current user medical data comprise user identification data, drug injection data and physical sign measurement data;

the historical data searching unit is used for acquiring a historical medical data set corresponding to the identification data if current user medical data uploaded by the medical terminal of the Internet of things is received; wherein the historical medical data set comprises an electronic medical record data subset, an examination report data subset, a medication record data subset and a historical treatment scheme data subset;

The current user input data acquisition unit is used for calling a preset first keyword set and a first extraction strategy corresponding to the first keyword set, extracting a target data set in a historical medical data set corresponding to the identification data, and combining the target data set with the current user medical data to obtain current user input data;

the classification result acquisition unit is used for forming a current user data input vector by semantic vectors corresponding to all fields in the current user input data, and inputting the current user data input vector into a pre-trained convolutional neural network model to obtain a classification result corresponding to the current user data input vector;

the data type judging unit is used for judging whether the treatment path data type corresponding to the classification result is larger than 1;

the standard vector acquisition unit is used for acquiring standard user data input vectors corresponding to the treatment path data types corresponding to the classification results if the treatment path data types corresponding to the classification results are greater than 1; and

and the optimal path data acquisition unit is used for calculating vector similarity between the standard user data input vector corresponding to the classification result and the current user data input vector respectively, and acquiring the standard user data input vector corresponding to the maximum vector similarity and the corresponding treatment path data as optimal treatment path data.

In a third aspect, an embodiment of the present invention further provides a computer device, which includes a memory, a processor, and a computer program stored in the memory and capable of running on the processor, where the processor implements the method for classifying user data based on internet of things data according to the first aspect when executing the computer program.

In a fourth aspect, an embodiment of the present invention further provides a computer readable storage medium, where the computer readable storage medium stores a computer program, where the computer program when executed by a processor causes the processor to perform the method for classifying user data based on internet of things data according to the first aspect.

The embodiment of the invention provides a user data classification method, a device, computer equipment and a storage medium based on internet of things data, which comprise the steps of judging whether current user medical data uploaded by an internet of things medical terminal are received or not; if current user medical data uploaded by the medical terminal of the Internet of things is received, acquiring a historical medical data set corresponding to the identification data; invoking a preset first keyword set and a first extraction strategy corresponding to the first keyword set, and extracting a target data set in a historical medical data set corresponding to the identification data so as to be combined with the current user medical data to obtain current user input data; the semantic vectors corresponding to the fields in the current user input data form a current user data input vector, the current user data input vector is input into a pre-trained convolutional neural network model, and a classification result corresponding to the current user data input vector is obtained; judging whether the treatment path data category corresponding to the classification result is greater than 1; if the treatment path data types corresponding to the classification results are larger than 1, standard user data input vectors corresponding to the treatment path data types corresponding to the classification results are obtained; and calculating vector similarity between the standard user data input vector and the current user data input vector, which correspond to the classification result respectively, and obtaining the standard user data input vector and the corresponding treatment path data, which correspond to the maximum vector similarity, as the optimal treatment path data. According to the method, the user medical data is quickly acquired based on the Internet of things equipment, the optimal treatment scheme data is quickly and accurately recommended based on the historical medical data, and the cost for acquiring the optimal treatment scheme data is reduced.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

Fig. 1 is an application scenario schematic diagram of a user data classification method based on internet of things data according to an embodiment of the present invention;

fig. 2 is a flow chart of a user data classification method based on internet of things data according to an embodiment of the present invention;

fig. 3 is a schematic block diagram of a user data classification device based on internet of things data according to an embodiment of the present invention;

fig. 4 is a schematic block diagram of a computer device according to an embodiment of the present invention.

Detailed Description

The following description of the embodiments of the present invention will be made clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are some, but not all embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.

It should be understood that the terms "comprises" and "comprising," when used in this specification and the appended claims, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.

It is also to be understood that the terminology used in the description of the invention herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used in this specification and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise.

It should be further understood that the term "and/or" as used in the present specification and the appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes such combinations.

Referring to fig. 1 and fig. 2, fig. 1 is a schematic application scenario diagram of a user data classification method based on internet of things data according to an embodiment of the present invention; fig. 2 is a flow chart of a user data classification method based on internet of things data, which is provided by the embodiment of the invention, and is applied to a cloud server, and the method is executed by application software installed in the cloud server.

As shown in fig. 2, the method includes steps S110 to S170.

S110, judging whether current user medical data uploaded by the medical terminal of the Internet of things are received or not; the current user medical data comprises user identification data, medicine injection data and physical sign measurement data.

In this embodiment, in order to understand the technical solution more clearly, a terminal related to a specific implementation scenario is described in detail.

Firstly, thing networking medical terminal can be the thing networking device such as glucometer, insulin syringe, thermometer when practicing, and these thing networking equipment is different with traditional glucometer, insulin syringe, thermometer, is provided with wireless communication module such as bluetooth low energy module, 4G 5G communication module on these thing networking equipment, through this transformation later, some parameter values that detect through thing networking medical terminal are: 1) The cloud server can be directly uploaded to the cloud server; 2) The method can be firstly interconnected with intelligent terminals such as a smart phone and the like, then the parameter values are uploaded to the intelligent terminals, and then the parameter values are uploaded to a server by the intelligent terminals. For example, a chronic patient (such as a patient suffering from hypertension, diabetes, cardiovascular and cerebrovascular diseases and the like) needs to use the internet of things equipment for measuring parameters such as blood sugar, body temperature and the like for a long time, or insulin is injected regularly and quantitatively, and at the moment, after the corresponding parameter value is detected through the internet of things medical terminal, the parameter value can be finally uploaded to a cloud server in a networking manner to serve as the parameter value for monitoring the chronic diseases.

And the intelligent terminal is used for receiving the current user medical data uploaded by the medical terminal of the internet of things bound with the intelligent terminal, and sending the current user medical data to the cloud server for storage.

And thirdly, the cloud server is used for receiving the current user medical data uploaded by the medical terminal of the Internet of things or the intelligent terminal, and storing the current user medical data after structural data conversion and the history data of the corresponding user. And similar case data and recommended treatment path data can be matched in the database based on current user medical data. Further, the cloud server can generate patient figures, patient medication data curves, medication risk condition predictions, and the like for data corresponding to each user.

When the user uses the medical terminal of the Internet of things to measure parameters such as blood sugar, body temperature and the like or inject insulin at regular time and quantity, the current medical data of the user can be accurately obtained. Because the medical terminal equipment of the internet of things needs to bind the user identification data when initializing in order to identify which user uses the medical terminal equipment of the internet of things to upload the user medical data.

For example, referring to the existing bluetooth connection process of the smart band and the smart phone and the initial binding process of the user identification data, after the bluetooth connection is performed between the medical terminal of the internet of things and the smart terminal, the historical user medical data stored in the medical terminal of the internet of things can be obtained on the user interaction interface of the smart terminal (generally, when the connection is initialized, the historical user medical data stored in the medical terminal of the internet of things is empty, that is, the data is not stored yet), and whether the medical terminal of the internet of things is bound with the medical terminal of the internet of things is prompted. Once the user selects to bind the medical terminal of the Internet of things with the intelligent terminal, the method is equivalent to establishing a mapping corresponding relation between the unique equipment identification code of the medical terminal of the Internet of things and the unique equipment identification code of the intelligent terminal. In addition, because the unique equipment identification code of the intelligent terminal is in a corresponding relation with the user identification data (for example, the user uses the micro-signal code to bind the unique equipment identification code of the intelligent terminal), when the medical terminal of the internet of things uploads the current medical data of the user to the cloud server, the corresponding current medical data of the user also uploads the user identification data besides the data such as the drug injection data (for example, the insulin injection quantity), the physical sign measurement data (for example, the blood sugar value and the body temperature value) and the like.

S120, if current user medical data uploaded by the medical terminal of the Internet of things are received, acquiring a historical medical data set corresponding to the identification data; the historical medical data set comprises an electronic medical record data subset, an examination report data subset, a medication record data subset and a historical treatment scheme data subset.

In this embodiment, after the cloud server receives the current user medical data uploaded by the medical terminal of the internet of things, it needs to know which user corresponds to the data, so as to determine to store the current user medical data into the data table corresponding to the user.

The cloud server has at least four data sources in the historical medical data set corresponding to the identification data, and the data sources are respectively an electronic medical record, an inspection report, historical medication data and a historical treatment scheme stored in a medical system of a hospital. Because the above data stored in the medical system are unstructured data, the cloud server needs to perform data cleaning and structured storage after acquiring the data from the electronic medical record, the inspection report, the historical medication data and the historical treatment in the medical system. If the cloud server does not receive the current user medical data uploaded by the medical terminal of the internet of things, the cloud server continues to return to the execution step S110 after waiting for the preset delay waiting time (e.g. 10S).

In one embodiment, step S120 further includes:

receiving an electronic medical record, an examination report, historical medication data and a historical treatment scheme uploaded by a medical system server;

extracting a corresponding electronic medical record data initial subset in the electronic medical record by calling a preset second keyword set and a second extraction strategy corresponding to the second keyword set;

extracting a corresponding initial subset of inspection report data in the inspection report by calling a preset third keyword set and a third extraction strategy corresponding to the third keyword set;

text recognition is carried out on the pictures corresponding to the historical medication data through an OCR text recognition model, recognition texts corresponding to the historical medication data are obtained, and an initial subset of medication record data corresponding to the historical medication data is extracted through a preset fourth keyword set and a fourth extraction strategy corresponding to the fourth keyword set;

and extracting the initial subset of the historical treatment scheme data corresponding to the historical treatment scheme through a preset fifth keyword set and a fifth extraction strategy corresponding to the fifth keyword set.

In this embodiment, since the medical system server stores the electronic medical records, the examination reports, the historical medication data, and the data of the historical treatment plan corresponding to a large number of patients, respectively.

The electronic medical record of the patient is patient-related information which is directly recorded by computer operation when a doctor examines the patient, and at least comprises a home page, a course record, an examination result, a doctor's advice, an operation record, a nursing record and the like.

The examination report of a patient is generally an examination report such as a blood test examination report or a CT examination report, and for example, the CT examination report of a patient generally includes information such as the name, sex, age, subject, hospital stay, bed number, examination site, date of registration, examination name, examination method, visual presentation, and reporting doctor of the patient.

The patient's historical medication data is typically written in a doctor's prescription, which the doctor can take and upload to the medical system server. The doctor prescription paper generally includes information such as patient's name, age, subject matter, clinical diagnosis, address/telephone, prescription bill, prescribing doctor, fitter, proofer, medicine price, date of prescribing, etc.

The patient's historical treatment regimen is typically recorded by the doctor directly in the medical system server, i.e., one treatment regimen is recorded for each patient visit.

Because the information in the electronic medical record is numerous, in order to select key information in the information to realize data screening and cleaning, a second keyword set and a second extraction strategy corresponding to the second keyword set can be preset at the moment.

For example, the keywords included in the second set of keywords include patient name, identification number, gender, age, region area of the patient, family history, course record, examination result, medical advice, operation record, and nursing record, and the second extraction strategy is to extract specific values of the keywords corresponding to the second set of keywords.

For example, the keywords included in the third keyword set include name, identification number, gender, age, subject, hospitalization number, bed number, examination location, registration date, examination name, examination method, image representation (which may also be understood as examination result), and reporting doctor, and the third extraction policy is to extract specific values of the keywords corresponding to the third keyword set.

Since the patient's historical medication data is generally written in doctor's prescription, and is written in latin or chinese, text recognition is performed on the picture corresponding to the historical medication data by the OCR text recognition model at this time, so as to obtain a recognition text corresponding to the historical medication data (the recognition text is a computer text that can be understood by a computer) as the recognition text corresponding to the historical medication data.

Then, for example, the keywords included in the set fourth keyword set include name, identification card number, age, subject, clinical diagnosis, address/telephone, drug list, prescribing doctor, fitter, proofreader, drug price and date of drug, and the fourth extraction strategy is to extract the specific values of the keywords corresponding to the fourth keyword set.

For example, the keywords included in the set fifth keyword set include name, identification card number, gender, age, subject, hospitalization number, treatment scheme flow, and attending doctor, and the fifth extraction strategy is to extract specific values of the keywords corresponding to the fifth keyword set.

Because the electronic medical record data subset, the inspection report data subset, the medication record data subset and the historical treatment plan data subset all comprise common fields of patient names, identification card numbers and sexes, at the moment, one piece of data with the same names, identification card numbers and sexes in the electronic medical record data subset, the inspection report data subset, the medication record data subset and the historical treatment plan data subset can be combined into the historical medical data set of the same patient.

After the processing, the data cleaning of the historical original medical data corresponding to the user is realized, so that the historical medical data initial set corresponding to each user is obtained, and the specific value of each field of each data subset in the historical medical data initial set corresponding to each user can be subjected to data structuring conversion; for example, the specific value of the field of the treatment plan flow in the initial subset of the historical treatment plan data corresponding to the historical treatment plan is a large text, and in order to convert the structural data, the key word can be extracted to obtain the key word combination of the field of the treatment plan flow, so as to be used as the structural data corresponding to the field of the treatment plan flow.

In an embodiment, after extracting the initial subset of the historical treatment plan data corresponding to the historical treatment plan through a preset fifth keyword set and a fifth extraction policy corresponding to the fifth keyword set, the method further includes:

and extracting keywords from the values of the fields in the initial subset of electronic medical record data, the initial subset of inspection report data, the initial subset of medication record data and the initial subset of historical treatment plan data to obtain the subset of electronic medical record data corresponding to the initial subset of electronic medical record data, the subset of inspection report data corresponding to the initial subset of inspection report data, the subset of medication record data corresponding to the initial subset of medication record data and the subset of historical treatment plan data corresponding to the initial subset of historical treatment plan data.

In this embodiment, for example, a large text description corresponding to a specific value of a field of a treatment plan flow in an initial subset of historical treatment plan data is first performed on the specific value of the field of the treatment plan flow by word segmentation based on a probability statistical word segmentation model, so as to obtain a word segmentation result corresponding to the specific value of the field of the treatment plan flow;

Then extracting keywords which do not exceed a preset ranking value in the word segmentation result through a word frequency-inverse text frequency index model to be used as structural data corresponding to a specific value of a field of a treatment scheme flow; the specific values of other fields are extracted by keywords, so that the data structuring conversion is realized.

S130, a preset first keyword set and a first extraction strategy corresponding to the first keyword set are called, and a target data set in a historical medical data set corresponding to the identification data is extracted so as to be combined with the current user medical data to obtain current user input data.

In this embodiment, since the current user medical data uploaded by the current user through the medical terminal of the internet of things generally only includes the field values of the user identification data, the drug injection data and the physical sign measurement data, which are quite structured, the current user medical data is not required to be subjected to structured data conversion at this time, and is directly combined with the current user medical data to obtain the current user input data.

In one embodiment, the first set of keywords and a first extraction policy corresponding to the first set of keywords are stored in a blockchain network. Likewise, the second keyword set and a second extraction strategy corresponding to the second keyword set are stored in a blockchain network; the third keyword set and a third extraction strategy corresponding to the third keyword set are stored in a blockchain network; the fourth keyword set and a fourth extraction strategy corresponding to the fourth keyword set are stored in a blockchain network; the fifth set of keywords and a fifth extraction policy corresponding to the fifth set of keywords are stored in a blockchain network.

In this embodiment, the keyword set and the extraction policy in the cloud server may be stored in the blockchain network. The cloud server may act as one of the blockchain link point devices in the blockchain network. Blockchains are novel application modes of computer technologies such as distributed data storage, point-to-point transmission, consensus mechanisms, encryption algorithms, and the like. The Blockchain (Blockchain), which is essentially a decentralised database, is a string of data blocks that are generated by cryptographic means in association, each data block containing a batch of information of network transactions for verifying the validity of the information (anti-counterfeiting) and generating the next block. The blockchain may include a blockchain underlying platform, a platform product services layer, and an application services layer.

In one embodiment, step S130 includes:

invoking the first keyword set and a first extraction strategy corresponding to the first keyword set, and acquiring a target data set corresponding to the identification data and respectively corresponding to each field in the first keyword set from a historical medical data set;

and combining the target data set with repeated fields in the current user medical data and de-duplicating the repeated fields to obtain the current user input data.

In this embodiment, however, there is generally a repeated field in the target data set in the current user medical data and the corresponding historical medical data set, for example, the user identification data is a repeated field in the target data set in the current user medical data and the corresponding historical medical data set, and this repeated field may be de-duplicated and one of them may be reserved.

For example, the current user medical data includes 3 fields of user identification data, drug injection data, and physical sign measurement data and specific field values thereof, and the target data set in the historical medical data set corresponding to the identification data includes 10 fields of patient name, user identification data (specifically, identification card number), region to which the patient belongs, family medical history, patient condition, examination result, and drug administration list and specific field values thereof, and since the field of user identification data is repeated, the current user medical data is merged with the current user medical data, and the obtained current user input data includes 10 fields of user identification data, drug injection data, physical sign measurement data, patient name, region to which the patient belongs, family medical history, patient condition, examination result, and drug administration list and specific field values thereof.

S140, forming a current user data input vector by semantic vectors corresponding to all fields in the current user input data, and inputting the current user data input vector into a pre-trained convolutional neural network model to obtain a classification result corresponding to the current user data input vector.

In this embodiment, for example, the obtained current user input data includes 10 fields including user identification data, drug injection data, sign measurement data, patient name, region to which the patient belongs, family history, patient condition, inspection result, and drug administration list, and specific field values thereof, and at this time, semantic vectors corresponding to the 10 fields may be formed according to word vectors corresponding to the specific values corresponding to the 10 fields (i.e., according to the 10 word vectors and weights corresponding to the 10 word vectors, respectively, the semantic vectors corresponding to the 10 fields are obtained).

And after semantic vectors corresponding to all fields in the current user input data are obtained to form the current user data input vector, calling a convolutional neural network model which is stored in a cloud server and is trained, and inputting the current user data input vector into a convolutional neural network model which is trained in advance, so that a classification result corresponding to the current user data input vector can be obtained.

For example, when training a convolutional neural network model to be trained in a cloud server, a training set is generally obtained first, where each piece of training data in the training set includes two parts: firstly, training set input data, which comprises 10 field specific field values (generally, the field specific values of the 10 fields are converted into keyword combinations, so that the field specific values of the 10 fields can be conveniently converted into corresponding word vectors to form semantic vectors, and the semantic vectors corresponding to the field specific values of the 10 fields form corresponding training set input vector data); and secondly, the training set outputs data, which comprises treatment path classification results, generally at least corresponds to one treatment path for each type of disease, so that the classification results are generally obtained by judging the current user data input vector, and the treatment path to which the disease belongs is determined.

In one embodiment, step S140 includes:

Acquiring semantic vectors corresponding to fields in the current user input data, and summing each semantic vector to obtain a current user data input vector corresponding to the current user input data;

invoking a prestored convolutional neural network model;

and taking the current user data input vector corresponding to the current user input data as the input of the convolutional neural network model to obtain a corresponding classification result.

In this embodiment, it is assumed that the semantic vectors corresponding to the fields in the current user input data are all one-dimensional row vectors of 1×300, and at this time, vector addition is performed on the semantic vectors corresponding to the fields in the current user input data, so as to obtain a current user data input vector corresponding to the current user input data. And then, taking the current user data input vector corresponding to the current user input data as the input of the convolutional neural network model to carry out classification operation, and obtaining a corresponding classification result.

And S150, judging whether the treatment path data category corresponding to the classification result is larger than 1.

In this embodiment, after the classification result is determined according to the current user input data, since each classification result corresponds to at least one type of treatment path data, the treatment path data generally includes a specific value of a treatment plan flow field. Therefore, after the classification result corresponding to the current user input data is determined, the treatment scheme flow corresponding to the current user data can be obtained. If the type of the treatment path data corresponding to the classification result is greater than 1, the optimal treatment path data can be recommended to the user corresponding to the current user input data.

And S160, if the treatment path data type corresponding to the classification result is greater than 1, acquiring standard user data input vectors corresponding to the treatment path data types corresponding to the classification result.

In this embodiment, when the convolutional neural network to be trained is trained by the training set, the general training set includes a plurality of kinds of training subsets, and one kind of treatment path data corresponding to each training subset is used as the training set output data. Thus, one piece of training data is selected from each training subset to serve as standard training set input data corresponding to the type of treatment path data.

For example, the training set input data corresponding to 1000 treatment path data includes 1000 training subsets, and one training set input data is selected in each of the 1000 training subsets to be used as the standard training set input data corresponding to each training subset. Because the standard training set input data corresponding to each training subset corresponds to one standard user data input vector.

And when the classification result is 3 kinds of treatment path data in 1000 kinds of treatment path data, acquiring standard user data input vectors corresponding to the 3 kinds of treatment path data.

And if the treatment path data type corresponding to the classification result is equal to 1, acquiring the treatment path data corresponding to the treatment path data type as the optimal treatment path data.

And S170, calculating vector similarity between the standard user data input vector corresponding to the classification result and the current user data input vector, and obtaining the standard user data input vector corresponding to the maximum vector similarity and the corresponding treatment path data as the optimal treatment path data.

In this embodiment, after the standard user data input vectors corresponding to the 3 types of treatment path data are obtained, if the vector similarity between the standard user data input vector corresponding to one type of treatment path data and the current user data input vector is the maximum, the treatment path data corresponding to the treatment path data is selected as the optimal treatment path data.

In calculating the similarity between vectors, algorithms such as euclidean distance (Eucledian Distance), cosine similarity (Cosine Similarity), minkowski distance (Minkowski Distance), and jaccard similarity coefficient (Jaccard Similarity) may be used.

And at the moment, the optimal treatment path data can be pushed to the intelligent terminal corresponding to the current user input data. Furthermore, in the cloud server, patient portraits, patient medication data curves, medication risk condition predictions and the like can be generated for data corresponding to each user, and the generation of the data can be based on a historical medical data set of the user.

In an embodiment, step S170 further includes:

extracting target portrait data sets corresponding to all users in the historical medical data set by calling a preset sixth keyword set and a sixth extraction strategy corresponding to the sixth keyword set;

screening keywords in the target portrait data set corresponding to each user according to the target portrait data set corresponding to each user and a pre-stored keyword screening strategy to obtain a target portrait data simplified set corresponding to each user;

and calling a pre-stored keyword conversion strategy to correspondingly convert the target portrait data simplified set corresponding to each user into user portrait data.

In this embodiment, for example, the keywords included in the set sixth keyword set include the region where the patient belongs to, family history, patient symptoms, and inspection results, and the sixth extraction policy is to extract specific values of the keywords corresponding to the sixth keyword set, and at this time, extract specific values of the keywords corresponding to the sixth keyword set to form the target portrait data set corresponding to each user in the historical medical data set.

Since the number of keywords corresponding to the target portrait data set corresponding to each user is still more, and some numerical parameters are possible or concrete, important keywords in the keywords can be screened and converted into corresponding labels.

The label conversion strategy corresponding to each keyword (the strategy of converting the keyword into the corresponding label) is set in the cloud server, and at the moment, the user portrait data corresponding to each user can be obtained after the label conversion of the keyword.

According to the method, the user medical data is quickly acquired based on the Internet of things equipment, the optimal treatment scheme data is quickly and accurately recommended based on the historical medical data, and the cost for acquiring the optimal treatment scheme data is reduced. The method belongs to the field of intelligent medical treatment, and the construction of the intelligent city can be promoted through the scheme.

The embodiment of the invention also provides a user data classification device based on the internet of things data, which is used for executing any embodiment of the user data classification method based on the internet of things data. Specifically, referring to fig. 3, fig. 3 is a schematic block diagram of a user data classification device based on internet of things data according to an embodiment of the present invention. The user data classification device 100 based on the internet of things data may be configured in a cloud server.

As shown in fig. 3, the user data classification device 100 based on internet of things data includes: a current data receiving unit 110, a history data searching unit 120, a current user input data acquiring unit 130, a classification result acquiring unit 140, a data category judging unit 150, a standard vector acquiring unit 160, and a best path data acquiring unit 170.

A current data receiving unit 110, configured to determine whether current user medical data uploaded by the medical terminal of the internet of things is received; the current user medical data comprises user identification data, medicine injection data and physical sign measurement data.

In this embodiment, when the user uses the medical terminal of the internet of things to measure parameters such as blood sugar and body temperature, or inject insulin at regular time and quantity, the current medical data of the user can be accurately obtained. Because the medical terminal equipment of the internet of things needs to bind the user identification data when initializing in order to identify which user uses the medical terminal equipment of the internet of things to upload the user medical data.

The historical data searching unit 120 is configured to obtain a historical medical data set corresponding to the identification data if current user medical data uploaded by the medical terminal of the internet of things is received; the historical medical data set comprises an electronic medical record data subset, an examination report data subset, a medication record data subset and a historical treatment scheme data subset.

The cloud server has at least four data sources in the historical medical data set corresponding to the identification data, and the data sources are respectively an electronic medical record, an inspection report, historical medication data and a historical treatment scheme stored in a medical system of a hospital. Because the above data stored in the medical system are unstructured data, the cloud server needs to perform data cleaning and structured storage after acquiring the data from the electronic medical record, the inspection report, the historical medication data and the historical treatment in the medical system. If the cloud server does not receive the current user medical data uploaded by the medical terminal of the Internet of things, after waiting for the preset delay waiting time (for example, 10 s), the step of judging whether the current user medical data uploaded by the medical terminal of the Internet of things is received is continuously carried out.

In an embodiment, the user data classifying apparatus 100 based on the internet of things data further includes:

the historical data uploading and receiving unit is used for receiving the electronic medical record, the examination report, the historical medication data and the historical treatment scheme uploaded by the medical system server;

the electronic medical record data initial subset obtaining unit is used for extracting a corresponding electronic medical record data initial subset in the electronic medical record by calling a preset second keyword set and a second extraction strategy corresponding to the second keyword set;

an inspection report data initial subset obtaining unit, configured to extract an inspection report data initial subset corresponding to the inspection report by calling a third preset keyword set and a third extraction policy corresponding to the third keyword set;

the medication record data initial subset obtaining unit is used for carrying out text recognition on pictures corresponding to the historical medication data through an OCR text recognition model to obtain recognition texts corresponding to the historical medication data, and extracting a medication record data initial subset corresponding to the historical medication data through a preset fourth keyword set and a fourth extraction strategy corresponding to the fourth keyword set;

The history treatment scheme data initial subset obtaining unit is used for extracting a history treatment scheme data initial subset corresponding to the history treatment scheme through a preset fifth keyword set and a fifth extraction strategy corresponding to the fifth keyword set.

In an embodiment, the user data classifying device 100 based on the internet of things data further includes:

and the data subset acquisition unit is used for extracting keywords from values of various fields in the electronic medical record data initial subset, the inspection report data initial subset, the medication record data initial subset and the historical treatment scheme data initial subset to obtain the electronic medical record data subset corresponding to the electronic medical record data initial subset, the inspection report data subset corresponding to the inspection report data initial subset, the medication record data subset corresponding to the medication record data initial subset and the historical treatment scheme data subset corresponding to the historical treatment scheme data initial subset.

The current user input data obtaining unit 130 is configured to invoke a preset first keyword set and a first extraction policy corresponding to the first keyword set, and extract a target data set in a historical medical data set corresponding to the identification data, so as to combine the target data set with the current user medical data to obtain current user input data.

In one embodiment, the current user input data acquisition unit 130 includes:

the target data set acquisition unit is used for calling the first keyword set and a first extraction strategy corresponding to the first keyword set, and acquiring a target data set corresponding to the identification data and respectively corresponding to each field in the first keyword set in a historical medical data set;

and the data deduplication unit is used for carrying out data combination and deduplication on repeated fields in the target data set and the current user medical data to obtain current user input data.

The classification result obtaining unit 140 is configured to form a current user data input vector from semantic vectors corresponding to fields in the current user input data, and input the current user data input vector to a convolutional neural network model trained in advance, so as to obtain a classification result corresponding to the current user data input vector.

In one embodiment, the classification result obtaining unit 140 includes:

The current user data input vector acquisition unit is used for acquiring semantic vectors corresponding to all fields in the current user input data, and summing each semantic vector to obtain a current user data input vector corresponding to the current user input data;

the model calling unit is used for calling a prestored convolutional neural network model;

and the vector calculation classification unit is used for taking the current user data input vector corresponding to the current user input data as the input of the convolutional neural network model to obtain a corresponding classification result.

And a data type judging unit 150, configured to judge whether the treatment path data type corresponding to the classification result is greater than 1.

The standard vector obtaining unit 160 is configured to obtain standard user data input vectors corresponding to the treatment path data types corresponding to the classification result, if the treatment path data type corresponding to the classification result is greater than 1.

The optimal path data obtaining unit 170 is configured to calculate vector similarity between the standard user data input vector and the current user data input vector, which respectively correspond to the classification result, and obtain the standard user data input vector and the corresponding treatment path data, which correspond to the maximum vector similarity, as optimal treatment path data.

the target portrait data set acquisition unit is used for extracting target portrait data sets corresponding to all users in the historical medical data set by calling a preset sixth keyword set and a sixth extraction strategy corresponding to the sixth keyword set;

the data set simplifying unit is used for screening keywords in the target portrait data sets corresponding to the users according to the target portrait data sets corresponding to the users and a pre-stored keyword screening strategy to obtain the target portrait data simplifying sets corresponding to the users;

And the user portrait data generating unit is used for calling a pre-stored keyword conversion strategy and correspondingly converting the target portrait data simplified set corresponding to each user into user portrait data.

The device realizes the rapid acquisition of the medical data of the user based on the Internet of things equipment, rapidly and accurately recommends the optimal treatment scheme data based on the historical medical data, and reduces the cost of acquiring the optimal treatment scheme data.

The above-described user data sorting apparatus based on internet of things data may be implemented in the form of a computer program which may be run on a computer device as shown in fig. 4.

Referring to fig. 4, fig. 4 is a schematic block diagram of a computer device according to an embodiment of the present invention. The computer device 500 is a server, and the server may be a stand-alone server or a server cluster formed by a plurality of servers.

With reference to FIG. 4, the computer device 500 includes a processor 502, memory, and a network interface 505, connected by a system bus 501, where the memory may include a non-volatile storage medium 503 and an internal memory 504.

The non-volatile storage medium 503 may store an operating system 5031 and a computer program 5032. The computer program 5032, when executed, may cause the processor 502 to perform a user data classification method based on internet of things data.

The processor 502 is used to provide computing and control capabilities to support the operation of the overall computer device 500.

The internal memory 504 provides an environment for the execution of a computer program 5032 in the non-volatile storage medium 503, which computer program 5032, when executed by the processor 502, causes the processor 502 to perform a user data classification method based on internet of things data.

The network interface 505 is used for network communication, such as providing for transmission of data information, etc. It will be appreciated by those skilled in the art that the architecture shown in fig. 4 is merely a block diagram of some of the architecture relevant to the present inventive arrangements and is not limiting of the computer device 500 to which the present inventive arrangements may be implemented, and that a particular computer device 500 may include more or fewer components than shown, or may combine some of the components, or have a different arrangement of components.

The processor 502 is configured to execute a computer program 5032 stored in a memory, so as to implement the user data classification method based on internet of things data disclosed in the embodiment of the present invention.

Those skilled in the art will appreciate that the embodiment of the computer device shown in fig. 4 is not limiting of the specific construction of the computer device, and in other embodiments, the computer device may include more or less components than those shown, or certain components may be combined, or a different arrangement of components. For example, in some embodiments, the computer device may include only a memory and a processor, and in such embodiments, the structure and function of the memory and the processor are consistent with the embodiment shown in fig. 4, and will not be described again.

It should be appreciated that in embodiments of the present invention, the processor 502 may be a central processing unit (Central Processing Unit, CPU), the processor 502 may also be other general purpose processors, digital signal processors (Digital Signal Processor, DSP), application specific integrated circuits (Application Specific Integrated Circuit, ASIC), off-the-shelf Programmable gate arrays (FPGAs) or other Programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or the like. Wherein the general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.

In another embodiment of the invention, a computer-readable storage medium is provided. The computer readable storage medium may be a non-volatile computer readable storage medium. The computer readable storage medium stores a computer program, wherein the computer program realizes the user data classification method based on the internet of things data disclosed by the embodiment of the invention when being executed by a processor.

It will be clearly understood by those skilled in the art that, for convenience and brevity of description, specific working procedures of the apparatus, device and unit described above may refer to corresponding procedures in the foregoing method embodiments, which are not repeated herein. Those of ordinary skill in the art will appreciate that the elements and algorithm steps described in connection with the embodiments disclosed herein may be embodied in electronic hardware, in computer software, or in a combination of the two, and that the elements and steps of the examples have been generally described in terms of function in the foregoing description to clearly illustrate the interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.

In the several embodiments provided by the present invention, it should be understood that the disclosed apparatus, device and method may be implemented in other manners. For example, the apparatus embodiments described above are merely illustrative, and for example, the division of the units is merely a logical function division, there may be another division manner in actual implementation, or units having the same function may be integrated into one unit, for example, multiple units or components may be combined or may be integrated into another system, or some features may be omitted, or not performed. In addition, the coupling or direct coupling or communication connection shown or discussed with each other may be an indirect coupling or communication connection via some interfaces, devices, or elements, or may be an electrical, mechanical, or other form of connection.

The units described as separate units may or may not be physically separate, and units shown as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the embodiment of the present invention.

In addition, each functional unit in the embodiments of the present invention may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units may be implemented in hardware or in software functional units.

The integrated units may be stored in a storage medium if implemented in the form of software functional units and sold or used as stand-alone products. Based on such understanding, the technical solution of the present invention is essentially or a part contributing to the prior art, or all or part of the technical solution may be embodied in the form of a software product stored in a storage medium, comprising several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a magnetic disk, an optical disk, or other various media capable of storing program codes.

While the invention has been described with reference to certain preferred embodiments, it will be understood by those skilled in the art that various changes and substitutions of equivalents may be made and equivalents will be apparent to those skilled in the art without departing from the scope of the invention. Therefore, the protection scope of the invention is subject to the protection scope of the claims.

Claims

1. The user data classification method based on the data of the Internet of things is characterized by comprising the following steps of:

extracting an initial subset of historical treatment scheme data corresponding to the historical treatment scheme through a preset fifth keyword set and a fifth extraction strategy corresponding to the fifth keyword set;

extracting keywords from values of fields in the initial subset of electronic medical record data, the initial subset of inspection report data, the initial subset of medication record data and the initial subset of historical treatment plan data to obtain a subset of electronic medical record data corresponding to the initial subset of electronic medical record data, a subset of inspection report data corresponding to the initial subset of inspection report data, a subset of medication record data corresponding to the initial subset of medication record data and a subset of historical treatment plan data corresponding to the initial subset of historical treatment plan data;

calculating vector similarity between the standard user data input vector and the current user data input vector, which correspond to the classification results respectively, and obtaining the standard user data input vector and the corresponding treatment path data, which correspond to the maximum vector similarity, as optimal treatment path data;

The extraction strategy is to extract specific values of the keywords corresponding to the corresponding keyword sets.

2. The method for classifying user data based on internet of things data according to claim 1, wherein the forming a current user data input vector from semantic vectors corresponding to fields in the current user input data, inputting the current user data input vector into a pre-trained convolutional neural network model, and obtaining a classification result corresponding to the current user data input vector, includes:

invoking a prestored convolutional neural network model;

3. The method for classifying user data based on internet of things data according to claim 1, wherein the invoking the preset first keyword set and the first extraction policy corresponding to the first keyword set, extracting the target data set in the historical medical data set corresponding to the identification data, so as to combine with the current user medical data to obtain the current user input data, includes:

4. The method for classifying user data based on internet of things data according to claim 1, wherein the step of calculating vector similarity between the standard user data input vector and the current user data input vector, which correspond to the classification result, and obtaining the standard user data input vector and the corresponding treatment path data, which correspond to the maximum vector similarity, as the optimal treatment path data, further comprises:

5. The method of claim 1, wherein the first set of keywords and a first extraction policy corresponding to the first set of keywords are stored in a blockchain network.

6. User data classification device based on thing networking data, characterized by includes:

a history treatment plan data initial subset obtaining unit, configured to extract a history treatment plan data initial subset corresponding to the history treatment plan through a fifth preset keyword set and a fifth extraction policy corresponding to the fifth keyword set;

The data subset obtaining unit is used for extracting keywords from values of various fields in the electronic medical record data initial subset, the inspection report data initial subset, the medication record data initial subset and the historical treatment plan data initial subset to obtain an electronic medical record data subset corresponding to the electronic medical record data initial subset, an inspection report data subset corresponding to the inspection report data initial subset, a medication record data subset corresponding to the medication record data initial subset and a historical treatment plan data subset corresponding to the historical treatment plan data initial subset;

the optimal path data acquisition unit is used for calculating vector similarity between the standard user data input vector corresponding to the classification result and the current user data input vector respectively, and acquiring the standard user data input vector corresponding to the maximum vector similarity and the corresponding treatment path data as optimal treatment path data;

7. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the method of classifying user data based on internet of things data as claimed in any one of claims 1 to 5 when the computer program is executed.

8. A computer-readable storage medium, characterized in that the computer-readable storage medium stores a computer program, which when executed by a processor causes the processor to perform the user data classification method based on internet of things data according to any one of claims 1 to 5.