A medical training platform management system based on intelligent cloud

Abstract

The present invention discloses a medical training platform management system based on intelligent cloud, which relates to the field of medical training technology and is used to solve the problem that most users who want to learn in depth only stay on the knowledge data after brief search and cannot deeply understand other relevant medical knowledge. The problem that leads to low learning efficiency of such users includes judging whether it is necessary to turn on intelligent recommendation according to the degree of interest of different users in learning medical knowledge and the degree of demand for additional medical knowledge; by detecting and analyzing the medical knowledge data and user usage data in the medical training software, evaluating whether the user of this software needs to turn on intelligent push, first collecting the user's learning data and search data, building a user medical knowledge interest evaluation model, and recording the user evaluation index, and then classifying the users, judging whether different users need to turn on intelligent recommendation when using this software, ensuring the user's learning environment and improving the user's learning efficiency of medical knowledge.

Description

Technical Field

[0001] The present invention relates to the field of medical training technology, and more specifically, to a medical training platform management system based on intelligent cloud. Background Art

[0002] Cloud computing provides elastic services for storage, computing and network resources that can be expanded on demand. For medical training platforms, this means that a large amount of educational content, including videos, simulation experiments, documents, etc., can be stored, and resources can be dynamically adjusted according to the number of users and access needs.

[0003] The prior art has the following deficiencies:

[0004] As the quality requirements of the nation are increasing year by year, the targets of medical training have shifted from medical students and professional job seekers to the general public. The traditional online medical training mechanism is only suitable for people with relatively rich professional knowledge. For ordinary people without knowledge reserves, the time and energy spent on entering the medical field through medical training is unimaginable. The traditional medical training mechanism only gives the corresponding knowledge content when retrieving medical therapy, medical knowledge and medical skills data, and will not recommend relevant information knowledge to users. This makes most users who want to learn in depth only stay on the knowledge data after a brief search, and cannot deeply understand other related medical knowledge. This leads to low learning efficiency for such users.

[0005] In view of the above problems, the present invention proposes a solution. Summary of the invention

[0006] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a medical training platform management system based on an intelligent cloud. Through a medical training platform management system based on an intelligent cloud, it is determined whether it is necessary to enable intelligent recommendation according to the degree of interest of different users in learning medical knowledge and the degree of need for additional medical knowledge, so as to solve the problems raised in the above-mentioned background technology.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] A medical training platform management system based on intelligent cloud includes a data acquisition module, a data processing module and a data storage module, and includes the following processing steps:

[0009] Step S1, detecting the user's device, collecting the learning data of medical knowledge on the user's device, and then collecting the user's search data;

[0010] Step S2, constructing a medical knowledge interest evaluation model based on the user device learning data and search data to evaluate the user's interest in medical knowledge;

[0011] Step S3, based on the different users screened out by the medical knowledge interest evaluation index, collect the number of clicks on recommended medical knowledge by users with high interest levels, collect the proportion of additional medical knowledge of users with low interest levels in the total search data, and determine the degree of demand for additional medical knowledge by different users;

[0012] Step S4, judging whether it is necessary to enable intelligent recommendation based on the interest level of different users in learning medical knowledge and the level of their need for additional medical knowledge.

[0013] In a preferred embodiment, in step S1, the learning data includes the relevant playing time of accessing medical videos, and the search data includes the number of times the user searches for different medical knowledge.

[0014] Accessing the relevant playback time of medical videos is done by applying to the user to detect the video playback software of the current mobile device, accessing the relevant playback time of medical videos through API calls, and obtaining learning data;

[0015] The number of different medical knowledge searches by the user is obtained by acquiring the text data input by the user in the search box through the value attribute of the corresponding UI element, thereby obtaining the number of different medical knowledge searches by the user.

[0016] In a preferred embodiment, in step S2, the sum of the relevant playback time of the accessed medical videos is divided by the standard playback time to obtain the quotient, the number of different medical knowledge searches is divided by the standard search number to obtain the quotient, and the user's medical knowledge interest evaluation index is calculated according to the polynomial regression model;

[0017] Among them, in step S2, the medical knowledge interest evaluation model is a polynomial regression model.

[0018] In a preferred embodiment, the medical knowledge interest evaluation index is compared with a preset interest threshold to determine the current user's interest in medical knowledge;

[0019] If the medical knowledge interest evaluation index is lower than the interest threshold, such users are marked as low-interest users.

[0020] If the medical knowledge interest evaluation index is higher than the interest threshold, such users will be marked as high-interest users, and intelligent recommendation content will be enabled for high-interest users.

[0021] In a preferred embodiment, the number of clicks on the recommended medical knowledge page by users with high interest is collected, and the recommended click ratio coefficient is obtained by dividing the obtained number of clicks on the recommended medical knowledge page by the standard number of clicks, and the data ratio of the additional knowledge searched by users with low interest to the total search data is collected to obtain the additional click ratio coefficient, and the additional medical knowledge demand degree of users with high interest and users with low interest is obtained by combining multiple decision trees respectively;

[0022] Among them, the degree of demand for additional medical knowledge is reflected by the additional medical knowledge demand index of users with high interest and the additional medical knowledge demand index of users with low interest.

[0023] In a preferred embodiment, in step S4, it is determined whether to recommend additional knowledge to users with high interest or users with low interest based on the medical knowledge interest evaluation index and the degree of demand for additional medical knowledge through the calculation result of the Sigmoid function.

[0024] The technical effects and advantages of the digital cultural tourism management system based on multi-source data analysis of the present invention are as follows:

[0025] The present invention detects and analyzes the medical knowledge data and user usage data in the medical training software to evaluate whether the users of the software need to enable smart push. First, the user's learning data and search data are collected to build a user medical knowledge interest evaluation model, and the user evaluation index is recorded. Then the users are classified to determine whether different users need to enable smart recommendations when using the software, thereby ensuring the user's learning environment and improving the user's learning efficiency of medical knowledge. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 This is a logic flow chart of a medical training platform management method based on intelligent cloud in the present invention;

[0027] Figure 2 This is a flowchart of the steps of a medical training platform management method based on intelligent cloud computing of the present invention. DETAILED DESCRIPTION

[0028] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0029] The present invention is based on the detection and analysis of medical knowledge data and user usage data in medical training software to evaluate whether the users of the software need to enable intelligent push. First, the user's learning data and search data are collected, a user medical knowledge interest evaluation model is constructed, and the user evaluation index is recorded. Then the users are classified to determine whether different users need to enable intelligent recommendation when using the software, thereby ensuring the user's learning environment and improving the user's learning efficiency of medical knowledge.

[0030] Example 1

[0031] like Figure 1 , Figure 2 As shown, step S1 is to detect the user device currently using the software, collect the learning data of medical knowledge on the user device, and then collect the search data of the user through the software.

[0032] Step S2, constructing a medical knowledge interest evaluation model based on the user device learning data and the software search data to evaluate the user's interest in medical knowledge.

[0033] Step S3, based on the different users screened out by the medical knowledge interest evaluation index, collect the number of clicks on the recommended medical knowledge by users with high interest levels, collect the proportion of additional medical knowledge of users with low interest levels in the total search data, and determine the degree of demand for additional medical knowledge by different users.

[0034] Step S4, judging whether it is necessary to enable intelligent recommendation based on the interest level of different users in learning medical knowledge and the level of their need for additional medical knowledge.

[0035] The specific implementation is as follows:

[0036] In the data acquisition module of step S1, the medical training software applies to the user for detection of the video playback software of the current mobile device, and the relevant playback time of medical videos is accessed through API calls to obtain learning data. For example, by monitoring the relevant playback time of medical videos with the keywords "medicine" and "drug", the total playback time is counted. The user's medical knowledge reserve is evaluated, and after the user uses this software for a period of time, the number of different medical knowledge searches by the user using this software is collected. By monitoring the user interface control, the text data entered by the user in the search box can be obtained through the value attribute of the corresponding UI element, thereby obtaining the number of different medical knowledge searches by the user.

[0037] It should be noted that the API involved in step S1 is a set of rules and tools that define how software components communicate with each other. API allows different software systems to exchange data, access functions and perform operations. UI elements are the basic building blocks of user interfaces, which are used to enable interaction between users and computer programs or systems. These elements allow users to perform specific operations, input data or navigate the system. These include different controls and tools such as text boxes, buttons and menus, depending on the design of the application or system.

[0038] In step S2, the sum of the relevant playback time of the accessed medical videos is divided by the standard playback time to obtain the video ratio coefficient and mark it as BF. The number of different medical knowledge searches is divided by the standard search number to obtain the additional click ratio coefficient and mark it as CF. The user's medical knowledge interest evaluation index is calculated according to the polynomial regression formula. The specific expression can be: Zs = αBF + βCF. In this formula, Z S Evaluate the user's medical knowledge interest index.

[0039] It should be noted that the standard playback time and the standard search times are set according to actual conditions, such as 300 hours and 20 times. The expressions involved in the above steps are not unique, and the specific values ​​of the proportional coefficients α and β involved in the formula are set by those skilled in the art according to specific conditions and are not limited here.

[0040] The video ratio coefficient BF represents the video ratio coefficient obtained by dividing the total duration of videos with the keywords "medicine" and "drug" in the software played on the user's current mobile device by the standard video duration. The user's video playback duration data can be obtained by calculating the sum of the playback time and closing time of the corresponding video accessed by the video server. Under the data of the relevant playback duration of standard access to medical videos, the larger the total playback duration of such videos by users, the greater the user's interest in additional medical knowledge, the larger the video ratio coefficient, and the greater the medical knowledge interest evaluation index. Conversely, the smaller the user's relevant playback duration of accessing medical videos, the less interest in learning additional medical knowledge, the smaller the video ratio coefficient, and the smaller the medical knowledge interest evaluation index.

[0041] The additional click ratio coefficient CF represents the additional click ratio coefficient obtained by dividing the number of times users search for keywords of different products using this software by the number of standard searches. User search data can be obtained by monitoring the text data input by the user through the user interface control. The more times users browse different medical knowledge, the greater the user's interest in learning different knowledge. The larger the additional click ratio coefficient, the greater the user's medical knowledge interest evaluation index, and vice versa.

[0042] It should be noted that when obtaining the corresponding parameters of learning data and search data, the summary data within a specific time period is used. For example, the learning data and search data within two weeks are collected, which means the total relevant playback time of users visiting medical videos within two weeks and the number of repetitions of medical knowledge. The specific data acquisition time period is set according to the actual situation and will not be analyzed in detail here.

[0043] The medical knowledge interest of customers is evaluated according to the medical knowledge interest evaluation index, and users with higher knowledge interest are screened out. The specific process is as follows:

[0044] The medical knowledge interest evaluation index Zs is compared with the preset interest threshold G to determine the current user's interest in medical knowledge.

[0045] If the medical knowledge interest evaluation index Zs is lower than the interest threshold G, the user has a greater interest in medical knowledge, and such users are marked as low-interest users.

[0046] If the medical knowledge interest evaluation index Zs is higher than the interest threshold G, the user has less interest in medical knowledge, and such users are marked as high-interest users.

[0047] The preset interest threshold G here is determined by the technical personnel according to the specific situation and is not analyzed.

[0048] In step S3, since different users have different interests in learning medical knowledge, and the time they open the software and the time they allocate for learning are also different, different users are classified and users with higher interests in medical knowledge are marked. Different users are classified and judged.

[0049] The specific analysis is as follows:

[0050] Users with high learning interest, i.e. marked users, are taken as sample one, and intelligent recommendation content is enabled for such users. Intelligent recommendation content recommends similar medical knowledge to users with high interest, including manufacturers, etc., when they search for medical knowledge. At the same time, the number of clicks on the recommended medical knowledge pages of users with high interest is collected through user interface control monitoring. The number of clicks on the recommended medical knowledge pages obtained is divided by the standard number of clicks to obtain the recommended click ratio coefficient g. Users with low learning interest, i.e. non-marked users, are taken as sample two, and the user interface control is monitored again to collect the data ratio of such users' searches for additional knowledge to the total search data to obtain the additional click ratio coefficient h.

[0051] For the above data, the random forest integration method is adopted to improve the accuracy of the data by combining multiple decision trees. The following are the specific steps of random forest:

[0052] First, we collect enough samples of recommended click ratio coefficients of labeled users and merge them into a data set labeled G. Then, we select a certain proportion of samples from the data set as a small sample data set labeled G. 1 , press G to convert the original data set 1 The ratio is divided into several small data sets G 2 , G 3 , G 4 Etc. This ensures that the samples used for each decision tree training are different. Use these small data sets as nodes to build a decision tree and randomly select a sample in the decision tree node. The specific steps are as follows:

[0053] In the small dataset G 1 Randomly select a recommended click ratio coefficient sample g 1 , in the decision tree G 2 Randomly select a recommended click ratio coefficient sample g 2 ,According to this method, a sample is randomly selected in each small sample data set. Random selection can improve the accuracy of decision trees.

[0054] Taking the average of these randomly selected samples, we get Select a different ratio to split the data set G, get a new small data set as a new node to rebuild the decision tree, and get the average value by randomly selecting samples. Repeat the operation to build multiple decision trees, form a random forest, and obtain some average data Etc., arrange these data from small to large and take out the median The recommended click ratio coefficient g of the marked user is calculated with the median The ratio is used as the additional medical knowledge demand index of the marked user and the data is recorded as C. When the index is less than 1, that is, If the index is greater than or equal to 1, then the user is judged to have low demand for additional medical knowledge. It is judged that the user has a high demand for additional medical knowledge.

[0055] For unlabeled users, the above random forest method is also used to calculate the median The ratio of the extra clicks of the non-labeled user search additional knowledge to the total search data is h and the median The ratio is used as the additional medical knowledge demand index of the unmarked user and the data is recorded and marked as D. Then the degree of demand for additional medical knowledge of the unmarked user is judged according to the judgment criteria in the above steps.

[0056] The recommended click ratio coefficient g mentioned in the above step is the ratio of the number of clicks on the medical knowledge page recommended by the marked user divided by the standard number of clicks. When the standard number of clicks is determined, the more clicks on the medical knowledge page recommended by the marked user, the larger the recommended click ratio coefficient g is, because It is fixed after calculation based on the collected data, so the larger the recommendation click ratio coefficient g, the larger the additional medical knowledge demand index C of the mark, that is, the higher the demand of the marked user for additional medical knowledge; the smaller the recommendation click ratio coefficient g, the smaller the additional medical knowledge demand index C of the mark, that is, the lower the demand of the marked user for additional medical knowledge.

[0057] The additional click ratio coefficient h is the ratio of non-labeled users’ searches for additional medical knowledge to the total search data. Additional medical knowledge refers to medical knowledge other than the learning content specified in medical training. It is fixed after calculation based on the collected data, so the larger the extra click ratio coefficient h is, the larger the non-marked additional medical knowledge demand index D is, that is, the higher the non-marked user's demand for additional medical knowledge is, and vice versa.

[0058] It should be noted that the user interface control monitoring can call the number of times users click on the recommended page and the search terms in the user's access log recorded on the server side of this software. It should be noted that obtaining user behavior data involves privacy issues, and necessary data needs to be provided before the user accesses the page. The standard number of clicks in the above steps is formulated according to actual conditions and is not limited here. The proportion of small data sets involved in the random forest and the number of decision trees built are set according to actual conditions, such as setting the ratio to 1:10 and building 5 decision trees.

[0059] In step S4, the obtained medical knowledge interest evaluation index Zs and additional medical knowledge demand indexes C and D are compared with the preset recommendation threshold through the calculation result of the Sigmoid function to determine whether additional knowledge is recommended for the current user. The formula can be: In the formula, T is used as the calculation result to compare with the preset recommended threshold. In addition, x=W in the formula 1 Z S +W 2 C+W 3 D, where x is the weighted sum of inputs, W 1 , W 2 and W 3is the weight index and is greater than 0. The specific value is set by professionals in this field and is not limited here. It should be noted that the marked user only marks the additional medical knowledge demand index C. When the marked user uses the above formula to calculate the result, the D value is set to 0. Similarly, when the non-marked user uses the above formula to calculate the result, the C value is set to 0. It can be seen from the above formula that when the weight index is determined, the larger the medical knowledge interest evaluation index, the larger the additional knowledge demand index, the larger the weighted sum x, that is, the larger the calculation result T. The easier it is to exceed the preset recommendation threshold.

[0060] Use the above formula to compare the calculated result T with the preset recommended threshold T K When the calculated result T is greater than the recommended threshold T K When the calculation result T is less than the recommended threshold T, the recommended additional knowledge function of the current user is turned on. K , turn off the recommended additional knowledge function for the current user.

[0061] It should be noted that the Sigmoid function is mainly used as an activation function in neural networks. It can be used to output the probability value of an event. The closer the output value is to 1, the greater the probability of the event. It can also be used as an output value compared with a preset recommendation threshold. By setting a recommendation threshold, when the output value is greater than the preset recommendation threshold, the event is judged to be established. When the output value is less than the preset recommendation threshold, the event is judged not to be established. In addition, the marked user does not have the medical knowledge demand index of the unmarked user, so it is set to 0 during calculation. Similarly, the unmarked user does not have the medical knowledge demand index of the marked user, and it is also set to 0 to facilitate entering a formula for calculation. The preset recommendation threshold T that appears in the above steps K It can be formulated according to actual conditions, such as 0.9, etc., and will not be repeated here.

[0062] In addition, the weighted sum formula set in the above steps is not unique.

[0063] The present invention can analyze the learning needs of different users more accurately and reasonably provide medical knowledge for different users to train and learn by comprehensively calculating the trigger conditions of various judgment factors, comparing and analyzing the threshold evaluation values ​​and adding them as feedback data to the secondary judgment conditions, breaking the problems of single data or complicated pages in previous medical training software, and improving the current situation that users with low learning interest or low learning efficiency are confused in knowledge memory due to these problems, and users with high learning interest or perfect learning methods lack additional medical knowledge. The learning efficiency of different users is greatly improved, and it is more convenient. At the same time, it can also understand the demand of different users for their own medical training, which is convenient for subsequent software to better manage user data and implement operation strategies.

[0064] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0065] Example 2

[0066] The present invention provides a medical training platform management system based on intelligent cloud, which also includes a data acquisition module, a data processing module and a data storage module to implement the method described in Example 1.

[0067] The data collection module is used to collect the learning data of medical knowledge on the user's device, and then collect the user's search data and other data.

[0068] The data processing module is used to recommend content to users based on the data collected by the data collection module.

[0069] The data storage module is used to store all data generated during the management system processing.

[0070] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments may be implemented in whole or in part in the form of a computer program product.

[0071] Those of ordinary skill in the art will appreciate that the modules and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0072] In addition, each functional module in each embodiment of the present application may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0073] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0074] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A medical training platform management system based on intelligent cloud, including a data acquisition module, a data processing module and a data storage module, characterized in that: The processing steps include: Step S1, detecting the user's device, collecting the learning data of medical knowledge on the user's device, and then collecting the user's search data; Step S2, constructing a medical knowledge interest evaluation model based on the user device learning data and search data to evaluate the user's interest in medical knowledge; Step S3, based on the different users screened out by the medical knowledge interest evaluation index, collect the number of clicks on recommended medical knowledge by users with high interest levels, collect the proportion of additional medical knowledge of users with low interest levels in the total search data, and determine the degree of demand for additional medical knowledge by different users; Step S4, judging whether it is necessary to enable intelligent recommendation according to the interest of different users in learning medical knowledge and the need for additional medical knowledge; In step S1, the learning data includes the relevant playback time of accessing medical videos, and the search data includes the number of times the user searches for different medical knowledge. Accessing the relevant playback time of medical videos is done by applying to the user to detect the video playback software of the current mobile device, accessing the relevant playback time of medical videos through API calls, and obtaining learning data; The number of different medical knowledge searches by the user is obtained by obtaining the text data entered by the user in the search box through the value attribute of the corresponding UI element, thereby obtaining the number of different medical knowledge searches by the user; In step S2, the sum of the relevant playback time of the accessed medical videos is divided by the standard playback time to obtain the quotient, the number of different medical knowledge searches is divided by the standard search number to obtain the quotient, and the user's medical knowledge interest evaluation index is calculated according to the polynomial regression model; Among them, in step S2, the medical knowledge interest evaluation model is a polynomial regression model.

2. According to claim 1, a medical training platform management system based on intelligent cloud is characterized by: Compare the medical knowledge interest evaluation index with the preset interest threshold to determine the current user's interest in medical knowledge: If the medical knowledge interest evaluation index is lower than the interest threshold, such users are marked as low-interest users; If the medical knowledge interest evaluation index is higher than the interest threshold, such users will be marked as high-interest users, and intelligent recommendation content will be enabled for high-interest users.

3. The intelligent cloud-based medical training platform management system according to claim 2 is characterized by: The number of clicks on the recommended medical knowledge page by users with high interest is collected, and the recommended click ratio coefficient is obtained by dividing the number of clicks on the recommended medical knowledge page by the standard number of clicks. The data ratio of the search for additional knowledge by users with low interest to the total search data is collected to obtain the additional click ratio coefficient, and the degree of demand for additional medical knowledge by users with high interest and users with low interest is obtained by combining multiple decision trees respectively. Among them, the degree of demand for additional medical knowledge is reflected by the additional medical knowledge demand index of users with high interest and the additional medical knowledge demand index of users with low interest.

4. The intelligent cloud-based medical training platform management system according to claim 3 is characterized by: In step S4, it is determined whether to recommend additional knowledge to users with high interest or users with low interest based on the medical knowledge interest evaluation index and the degree of demand for additional medical knowledge through the calculation result of the Sigmoid function.

Images