Causally driven dynamic organization method of teaching resources under teaching and learning behavior data support

Abstract

The application discloses a kind of teaching and learning behavior data support under causally driven dynamic organization method and system of teaching resources, and the method comprises the following steps: obtaining online teaching and learning behavior data, the online teaching and learning behavior data include teaching activity data, student learning data and online education resource information;According to the online teaching and learning behavior data, resource access dynamic time sequence causality is modeled, and the causality information of online education resource access is determined;According to the causality information, the online education resource access heat is predicted by neural network;According to the online education resource access heat obtained by prediction, the file level teaching and learning behavior data support under causally driven dynamic organization process of teaching resources is constructed, to realize the causally driven dynamic organization of teaching resources under teaching and learning behavior data support.The application can effectively improve the education user data access efficiency, and can be widely applied in the field of computer technology.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer, in particular to a teaching resource dynamic organization method driven by causality under teaching and learning behavior data support. BACKGROUND

[0002] Education resource organization optimization has always been one of the research hotspots in education data storage. At present, most of the researches are based on traditional metadata rule reasoning and other methods for organization, which lacks scheduling flexibility for complex cross-time and space education resource access demand, seriously affecting the user learning experience. In recent years, resource organization optimization algorithms based on machine learning have made certain progress, and the demand flexibility has been greatly improved compared with traditional methods, but limited by the defects of machine learning in causal relationship learning, the effect of these resource organization methods in the environment of complex cross-time and space access demand of online education is very limited. SUMMARY

[0003] Therefore, the embodiment of the present application provides a teaching resource dynamic organization method driven by causality under teaching and learning behavior data support, which can effectively improve the education user data access efficiency.

[0004] An aspect of the embodiment of the present application provides a teaching resource dynamic organization method driven by causality under teaching and learning behavior data support, comprising:

[0005] Obtaining online teaching and learning behavior data, the online teaching and learning behavior data comprising teaching activity data, student learning data and online education resource information;

[0006] Modeling the dynamic time sequence causality of resource access according to the online teaching and learning behavior data, and determining the causal information of online education resource access;

[0007] According to the causal information, predicting the online education resource access heat through a neural network;

[0008] According to the predicted online education resource access heat, constructing a file-level teaching resource dynamic organization process driven by causality under teaching and learning behavior data support, so as to realize the teaching resource dynamic organization driven by causality under teaching and learning behavior data support.

[0009] Optionally, the obtaining online teaching and learning behavior data comprises:

[0010] Collecting course outline and time information in the service area as teaching activity data;

[0011] Collecting student class information and network behavior information in the service area as student learning data;

[0012] Collecting the inherent attribute information and accessed record information of online education resources as online education resource information;

[0013] According to the target task scene, target features are selected from the teaching activity data, student learning data and online education resource information, and observable multi-dimensional time series are assembled as source information for online education resource access feature analysis.

[0014] Optionally, the online teaching and learning behavior data is used to model resource access dynamic time series causality, and determine the causality information of online education resource access, including:

[0015] According to the probability distribution-based partitioner, the time period of local causal stability of resource access is partitioned to obtain online resource access dynamic causal partition information;

[0016] According to the online resource access dynamic causal partition information, an online resource access dynamic causal description is constructed;

[0017] According to the online resource access dynamic causal partition information and the online resource access dynamic causal description, online resource access dynamic causal discovery is performed to determine the causality information of online education resource access.

[0018] Optionally, the causality information is used to predict the online education resource access popularity through a neural network, including:

[0019] A dynamic causal structure set is obtained, and a causal neural network layer for online education resource access popularity prediction network is constructed;

[0020] According to the causal neural network layer, an online resource access popularity prediction model is constructed;

[0021] The online education resource access popularity is predicted through the online resource access popularity prediction model.

[0022] Optionally, the causal neural network layer is used to construct an online resource access popularity prediction model, including:

[0023] A time series message passing neural network is combined to construct an online education resource access process representation data information propagation update network;

[0024] Based on the propagation mechanism Mask method, a dynamic causal representation learning mechanism under the time series encoding-decoding architecture is constructed to learn the causal representation;

[0025] According to the learned causal representation, an online teaching and learning behavior state evolution network and an online resource access reaction state network are constructed to obtain the online education resource access state;

[0026] A time series differential network is used to design an online resource access popularity predictor based on a time series differential neural network through the online education resource access state.

[0027] Optionally, the online education resource access heat degree is predicted to construct a file-level teaching and learning behavior data supported causal driving teaching resource dynamic organization process, including:

[0028] Obtain the data storage resource space situation, network load topology situation and each data node system load situation in each service area;

[0029] According to the predicted online education resource access heat degree, the online education resource access situation in each service area is obtained;

[0030] According to the network load topology situation and the online education resource access situation in each service area, the network load situation of each service in the online education system in a future period of time is estimated;

[0031] According to the online education resource access situation in each service area, the number of online education resource file copies in each service area is determined according to the predicted access heat degree;

[0032] According to the data storage resource space situation in each service area, the data node system load situation, the number of online education resource file copies in each service area and the network load situation of each service in the online education system in a future period of time, the execution resource consumption situation of the online education resource file copy adjustment system is calculated;

[0033] According to the number of online education resource file copies and the execution resource consumption situation, the final number of online education resource file copies in each service area is calculated, and corresponding file copy generation and deletion update operations are performed to complete the online education resource organization adjustment.

[0034] Another aspect of the embodiment of the application also provides a teaching and learning behavior data supported causal driving teaching resource dynamic organization system, including:

[0035] The first module is used for obtaining online teaching and learning behavior data, and the online teaching and learning behavior data includes teaching activity data, student learning data and online education resource information;

[0036] The second module is used for modeling resource access dynamic time sequence causality according to the online teaching and learning behavior data, and determining causal information of online education resource access;

[0037] The third module is used for predicting online education resource access heat degree through a neural network according to the causal information;

[0038] The fourth module is configured to construct a file-level teaching and learning behavior data supported causal driving teaching resource dynamic organization process according to the predicted online education resource access heat, so as to realize the teaching and learning behavior data supported causal driving teaching resource dynamic organization.

[0039] Another aspect of the embodiment of the present application further provides an electronic device, comprising a processor and a memory.

[0040] The memory is configured to store a program.

[0041] The processor executes the program to realize the method as described above.

[0042] Another aspect of the embodiment of the present application further provides a computer readable storage medium, which stores a program, and the program is executed by a processor to realize the method as described above.

[0043] The embodiment of the present application further discloses a computer program product or a computer program, which comprises computer instructions stored in a computer readable storage medium. A processor of a computer device can read the computer instructions from the computer readable storage medium, and the processor executes the computer instructions, so that the computer device executes the method as described above.

[0044] The embodiment of the present application acquires online teaching and learning behavior data, wherein the online teaching and learning behavior data comprises teaching activity data, student learning data and online education resource information; according to the online teaching and learning behavior data, resource access dynamic time sequence causality is modeled, and causal information of online education resource access is determined; according to the causal information, online education resource access heat is predicted through a neural network; according to the predicted online education resource access heat, a file-level teaching and learning behavior data supported causal driving teaching resource dynamic organization process is constructed, so as to realize the teaching and learning behavior data supported causal driving teaching resource dynamic organization. The present application can effectively improve the education user data access efficiency. BRIEF DESCRIPTION OF DRAWINGS

[0045] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0046] Figure 1 The overall step flowchart provided by the embodiment of the present application is provided. DETAILED DESCRIPTION

[0047] In order to make the purposes, technical solutions and advantages of the present application clearer, further detailed description will be given to the present application in combination with the drawings and examples. It should be understood that the specific examples described herein are only used to explain the present application, and are not used to limit the present application.

[0048] In order to solve the problems in the prior art, the embodiment of the present application provides a teaching and learning behavior data supported causal driving dynamic organization method of teaching resources, comprising:

[0049] Obtaining online teaching and learning behavior data, the online teaching and learning behavior data comprising teaching activity data, student learning data and online education resource information;

[0050] Modeling the dynamic time sequence causality of resource access according to the online teaching and learning behavior data, and determining the causal information of online education resource access;

[0051] According to the causal information, predicting the online education resource access heat through a neural network;

[0052] According to the predicted online education resource access heat, constructing a file-level teaching and learning behavior data supported causal driving dynamic organization process of teaching resources, so as to realize the teaching and learning behavior data supported causal driving dynamic organization of teaching resources.

[0053] Optionally, the obtaining online teaching and learning behavior data comprises:

[0054] Collecting course outline and time information in a service area as teaching activity data;

[0055] Collecting student class information and network behavior information in the service area as student learning data;

[0056] Collecting the intrinsic attribute information and accessed record information of online education resources as online education resource information;

[0057] According to a target task scenario, selecting target features from the teaching activity data, student learning data and online education resource information, and assembling observable multi-dimensional time series as source information for online education resource access feature analysis.

[0058] Optionally, the modeling the dynamic time sequence causality of resource access according to the online teaching and learning behavior data, and determining the causal information of online education resource access comprises:

[0059] According to a probability distribution based divider, the time period of local causal stability of resource access is divided, and online resource access dynamic causal boundary information is obtained;

[0060] According to the online resource access dynamic causal boundary information, constructing online resource access dynamic causal description;

[0061] According to the online resource access dynamic causal boundary information and the online resource access dynamic causal description, online resource access dynamic causal discovery is performed to determine the causal information of online education resource access.

[0062] Optionally, the online education resource access heat is predicted by a neural network according to the causal information, comprising:

[0063] A set of dynamic causal structures is obtained, and a causal neural network layer oriented to online education resource access heat prediction network is constructed;

[0064] According to the causal neural network layer, an online resource access heat prediction model is constructed;

[0065] The online education resource access heat is predicted by the online resource access heat prediction model.

[0066] Optionally, the online resource access heat prediction model is constructed according to the causal neural network layer, comprising:

[0067] Combined with a time sequence message passing neural network, an online education resource access process representation data information propagation update network is constructed;

[0068] Based on a propagation mechanism Mask method, a dynamic causal representation learning mechanism under a time sequence encoding-decoding architecture is constructed to learn causal representation;

[0069] According to the learned causal representation, an online teaching and learning behavior state evolution network and an online resource access reaction state network are constructed to obtain online education resource access state;

[0070] A time sequence differential network is used to design an online resource access heat predictor based on a time sequence differential neural network through online education resource access state.

[0071] Optionally, the online education resource access heat predicted is used to construct a causal driven teaching resource dynamic organization process under the support of file level teaching and learning behavior data, comprising:

[0072] The data storage resource space situation, network load topology situation, and each data node system load situation in each service area are obtained;

[0073] According to the predicted online education resource access heat, the online education resource access situation in each service area is obtained;

[0074] According to the network load topology situation and the online education resource access situation in each service area, the network load situation of each service in the online education system in a future period of time is estimated;

[0075] According to the online education resource access situation in each service area, the number of online education resource file copies in each service area is determined according to the predicted access heat size.

[0076] According to the data storage resource space situation in each service area, the system load situation of each data node, the number of online education resource file copies in each service area and the network load situation of each service in the online education system in the future period of time, the execution resource consumption situation of the online education resource file copy adjustment system is calculated.

[0077] According to the number of online education resource file copies and the execution resource consumption situation, the final number of online education resource file copies in each service area is calculated, and corresponding file copy generation and deletion update operations are performed to complete the online education resource organization adjustment.

[0078] Another aspect of the embodiment of the application also provides a teaching and learning behavior data supported causal driven teaching resource dynamic organization system, comprising:

[0079] A first module is used for obtaining online teaching and learning behavior data, and the online teaching and learning behavior data comprises teaching activity data, student learning data and online education resource information.

[0080] A second module is used for modeling resource access dynamic time sequence causality according to the online teaching and learning behavior data, and determining causal information of online education resource access.

[0081] A third module is used for predicting online education resource access heat through a neural network according to the causal information.

[0082] A fourth module is used for constructing a file level teaching and learning behavior data supported causal driven teaching resource dynamic organization process according to the predicted online education resource access heat, so as to realize the teaching and learning behavior data supported causal driven teaching resource dynamic organization.

[0083] Another aspect of the embodiment of the application also provides an electronic device comprising a processor and a memory.

[0084] The memory is used for storing a program.

[0085] The processor executes the program to realize the method as described above.

[0086] Another aspect of the embodiment of the application also provides a computer readable storage medium, wherein the storage medium stores a program, and the program is executed by a processor to realize the method as described above.

[0087] This invention also discloses a computer program product or computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device can read the computer instructions from the computer-readable storage medium and execute the computer instructions, causing the computer device to perform the aforementioned method.

[0088] The specific implementation process of the present invention will now be described in detail with reference to the accompanying drawings:

[0089] like Figure 1 As shown, this invention proposes a causal-driven dynamic organization method for teaching resources supported by teaching and learning behavior data. By analyzing historical data of teaching and learning behaviors, it discovers potential dynamic temporal causal relationships in accessing educational resources and designs a causal-driven dynamic organization method for teaching resources supported by teaching and learning behavior data, thereby effectively improving the data access efficiency of educational users.

[0090] The implementation process will be explained in detail below:

[0091] 1. Online teaching and learning behavior data collection

[0092] This invention proposes a causal-driven dynamic organization method for teaching resources supported by teaching and learning behavior data, applicable to geographically isolated online education systems. The entire system consists of multiple geographically non-overlapping service areas {…,P i Information data collection is conducted locally, on a service area basis, while educational resource organization is carried out globally across the entire system. Online teaching and learning behavior data collection mainly includes teaching activity data collection, student learning data collection, and online educational resource information collection. Within a specific service area, the collection of teaching activity data mainly includes course outlines and time information, which is represented by {…,C}. t ,…}, where C t =(…,c it ,…)},c it This represents the value of the i-th field in the specific teaching activity information. Student learning data collection within a certain service area mainly includes student class attendance information and online behavior information, etc. This part of the data is represented by {…,S}. t ,…}, where S t =(…,s it ,…)},s it This represents the value of the i-th field in the specific student learning behavior information. Online educational resource information mainly includes the inherent attributes of the online educational resources and access records, etc. This part of the data is represented by {…,R}. t ,…}, where R t =(…,r it,…)}, r it represents the value of the i-th field of the specific education information. t t t , etc. are observable data, and specific features can be selected from them to form an observable multi-dimensional time series X according to specific task scenarios, as the source information for specific online education resource access feature analysis.

[0093] 2. Dynamic timing dynamic causal modeling of resource access

[0094] The access of online education resources has significant cross-time and space characteristics, and there are great differences in the access of online education resources in different regions and time periods. The internal causal relationship of resource access will dynamically change. The present application models the dynamic timing causal relationship of resource access, and provides information for subsequent teaching and learning behavior data supported causal-driven dynamic organization of teaching resources.

[0095] In order to obtain the characteristics of online education resource access in different time periods, the problem of time period division for rule discovery needs to be solved first. The present application proposes an online resource dynamic causal modeling method, which mainly assumes that the online resource access features are stable in a period of time, i.e. local causal stability, but the online resource access features are dynamically changing in a long period of time, i.e. global resource causal dynamics. The modeling process mainly includes the following aspects.

[0096] 1) Dynamic causal demarcation of online resource access: This part is to demarcate and divide the time period of local causal stability of resource access. The present application proposes a demarcator D based on probability distribution, p(d t | l t ) represents the specific implementation of the demarcator, and l t represents the local stable causal structure for which the demarcator needs to be divided into a continuous time period. The selection of probability distribution can be determined according to specific task scenarios. The present application takes Bernoulli probability distribution as an example:

[0097]

[0098] Among them, condition 1 is count(d ＜t ) ≥ Max N (the symbol subscript <t represents within t time) indicates that the collected teaching and learning behavior observation sub-time series exceeds the specified maximum number, and condition 2 is count(d ＜t ) ≤ Minx L indicates that the length of the collected teaching and learning behavior observation sub-time series reaches the specified minimum number, represents the probability distribution parameter fitted by the neural network.

[0099] ​​2), online resource access dynamic causal description: combined with equation (1), the entire process of online education resource access feature local static global dynamic can be described as:

[0100]

[0101] Wherein, the global dynamic causal conversion G of g t The specific implementation can be described as follows:

[0102]

[0103] Wherein, h t Indicates that the time sequence neural network is used for g <t Encoding result, described as follows:

[0104]

[0105] Local causal conversion is similar to global causal conversion, described as follows:

[0106]

[0107] Wherein, Described as follows:

[0108]

[0109] The equation (2) of the application is described as follows:

[0110] log p(X)=logΣ D ∫ G,L p(X,G,L,D), (7)

[0111] Equation (6) can be approximated and optimized by the method of evidence lower bound (ELBO), that is,

[0112]

[0113] The decomposition form of equation (8) is expanded as follows:

[0114] q(G,L,D|X)=q(D|X)q(G|D,X)q(L|G,D,X), (9)

[0115] It can be seen that the demarcator D is independent of G, D, and q(D|X) can be decomposed as:

[0116]

[0117] The global dynamic causal predictor of equation (9) where q(z t The specific implementation of |D,X) is as follows:

[0118]

[0119] where, is the forward and backward feature representation encoding of the observable online education resource access related time series model.

[0120] 3) Online resource access dynamic causal discovery: with the help of the above causal boundary and causal description related hypothesis and calculation, the present application performs online resource access dynamic causal discovery on the observable multidimensional time series X, providing an information foundation for subsequent online education resource access heat prediction. The causal model described by the present application is as follows:

[0121]

[0122] where, pa i represents the cause of the observed variable x i , τ i represents a smooth function affected by unobserved confounding factors, and ε i is the interference term, is the causal function of the kth local stable causality. It is assumed that X k (t) = (x1(t), x2(t), …) is a multidimensional observed time series of teaching and learning, resource access, and X k The internal causal relationship is described as follows:

[0123]

[0124] wherein the present application assumes that f j is the prior, and τ i is μ, V are the mean and variance of the Gaussian process. To simplify the estimation of the local stable causal structure of online education resource access, it can be assumed that the noise of equation (13) is a Gaussian random distribution variable. Local stable causal structure estimation can be performed by maximizing the marginal likelihood probability, and the objective function is:

[0125]

[0126] where y k = N x ·{f j (t), g i (t)} T + [{ε i (t)} i,t ] T , m is N x μ, N is the kth local stable causal structure.

[0127] 3. Resource access heat prediction based on causal machine learning:

[0128] 1. Dynamic causal neural network module design: the present application predicts online education resource access heat through a neural network, which needs to use causal information to overcome the defects of associative learning. This process needs to construct a causal structure embedding network model. First, combine the above-mentioned dynamic causal structure set N, and use the differentiable programming technology to construct a causal neural network layer for online education resource access heat prediction network. This dynamic time series causal network layer can be described as:

[0129]

[0130] where A t is a representation of a directed acyclic graph adjacency matrix at time t, is a set of variables subject to Gaussian distribution, z t represents the causal representation learned based on the causal graph model at time t. Then, with the help of this structure, the intervention operation mechanism of the causal representation network is designed. The present application initially realizes the whole process of describing the generation of each sub-variable from the corresponding parent variable in the resource access local causal network through a simple mask mechanism. The adjacency matrix A t is split into a subset of n resource access influencing factors, i.e. The processing mechanism can be represented as:

[0131]

[0132] where represents element-wise multiplication, is the i-th subset of the adjacency matrix A t and is used to represent the causal correlation strength of the local stable causal representation z t , is a differentiable nonlinear function, represents the learnable parameters of , and the hidden variable represents that all non-parent teaching and learning behavior variables have been masked, i.e. only contains the parent information of . Since the mask mechanism simulates the intervention operation, it can effectively help the neural network learn causal information. Finally, combined with the evidence lower bound reasoning learning strategy, the time series loss function of the potential resource access heat calculation task is constructed, the causal network module integration design for the resource access dynamic evolution process in the continuous time domain is completed, and the resource access time series causal bias embedding is realized.

[0133] 2) Online resource access heat prediction based on causal neural network: based on the above dynamic causal network module and online resource access heat prediction demand, the application designs an online resource access heat prediction model based on causal neural network. First, combined with the time series message passing neural network, the online education resource access process representation data information propagation update network is constructed, which solves the problem of irregular input multi-modal features of the prediction model, and the information update propagation calculation is shown in equation (17):

[0134]

[0135] Where, S i (t - ) is the state vector of the source node data feature i at time t, S j (t - ) is the state vector of the destination node data feature j, τ represents the same or different modal feature propagation type identifier, msg τ is a learnable information propagation update function. Then based on the propagation mechanism Mask method of equation (17) above, the dynamic causal representation learning mechanism under the time series encoding-decoding architecture is constructed, and based on the learned causal representation, the online teaching and learning behavior state evolution network and the online resource access reaction state network are constructed, which are used to obtain the online education resource access state. Finally, using the time series differential network, an online resource access heat predictor based on the time series differential neural network is designed, which is formally represented as follows:

[0136] H t = OEDNET (H t-1 , f, f b , f p , θ b , θ p , θ, t begin , t end ) (18)

[0137] Where, H t represents the online resource access causal structure hidden state, f b represents the online teaching and learning behavior state evolution network, f p represents the online resource access reaction state network, f represents the resource access heat prediction network, θ * represents the model parameters, which can be realized by ordinary time series neural network, and the online resource access heat predictor of equation (18) can realize the online resource access heat prediction at any time resolution in the future.

[0138] 4) Causally driven teaching resource dynamic organization process supported by teaching and learning behavior data

[0139] The application is based on the online education resource heat prediction result, and proposes a file-level teaching and learning behavior data supported causal driving teaching resource dynamic organization process, mainly including the following main steps:

[0140] The first step is to obtain data storage resource space conditions (denoted as LST) in each service area, network load topology conditions (denoted as NLT), and system load conditions (denoted as SLT) of each data node;

[0141] The second step is to obtain online education resource access conditions (denoted as FHT) in each service area by using the online education resource access heat prediction method;

[0142] The third step is to combine the NLT obtained in the first step, utilize the FHT, and estimate network load conditions (denoted as NLT') of the online education system in each service area in a future period of time by using a method commonly used in the field;

[0143] The fourth step is to determine the number of online education resource file copies (denoted as EPT) in each service area according to the predicted access heat by using a simple statistical method according to the FHT;

[0144] The fifth step is to calculate the resource consumption conditions (denoted as EIT) of the online education resource file copy adjustment system by using LST, SLT, EPT and NLT', and the resource consumption conditions can be obtained by using a method commonly used in the field, such as the simplest linear calculation;

[0145] The sixth step is to combine EIT and EPT, calculate the final number of online education resource file copies in each service area, and perform corresponding file copy generation and deletion update operations to complete the online education resource organization adjustment in this round.

[0146] The above six steps are sequentially run according to the set estimation time period, and the teaching and learning behavior data supported causal driving teaching resource dynamic organization is realized.

[0147] In summary, the application proposes a teaching and learning behavior data supported causal driving teaching resource dynamic organization method, potential education resource access dynamic time sequence causal correlation is found by analyzing teaching and learning behavior historical data, and a teaching and learning behavior data supported causal driving teaching resource dynamic organization method is designed, so that the education user data access efficiency is effectively improved.

[0148] In some alternative embodiments, the function / operations mentioned in the block diagrams can not occur in the order mentioned in the operational illustrations. For example, depending on the involved function / operation, two blocks shown in succession can in fact be executed substantially concurrently or the blocks can sometimes be executed in reverse order, depending upon the functionality / operations involved. Furthermore, embodiments presented and described in the flowcharts are only examples of implementing the present application. Alternative embodiments are possible where some of the steps are omitted, wherein additional steps are added, or wherein some of the steps are performed in a different order. It should be understood that the order of steps presented and described in the flowcharts illustrates implementations of the present application. The steps presented and described in the flowcharts are not necessarily performed in the order presented and described. Steps from one exemplary flowchart can be performed in a different order.

[0149] Furthermore, although the present application has been described in the context of functional modules, it is to be understood that one or more of the functions and / or features described can be integrated in a single physical device and / or software module, or one or more functions and / or features can be implemented in separate physical devices or software modules. It will also be appreciated that detailed discussion of the actual implementation of each module is not necessary for an understanding of the present application. Rather, the actual implementation of the modules, in light of the description of the properties, functions and interrelationships of the various functional modules disclosed herein, will be apparent to one of ordinary skill in the art. Thus, the present application is not limited to the embodiments described herein but instead is limited only by the claims appended hereto. It will also be appreciated that the particular conceptual terminology employed herein is for purposes of clarity and description and is not intended to limit the scope of the present application. Rather, the specific conceptual terminology employed is for the purpose of description.

[0150] If the functions are implemented in the form of software function units and sold or used as independent products, they can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application essentially or the part of the prior art that makes a contribution or the part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in the embodiments of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, and various media that can store program codes.

[0151] The logic and / or steps represented in the flowcharts and / or described herein, for example, can be embodied in non-transitory computer-readable media, executed by one or more computing devices, and / or in any other way. The logic and / or steps represented in the flowcharts and / or described herein, for example, can be considered a list of executable instructions for implementing logic functions, and can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, processor- containing system, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions. For purposes of this specification, a "computer-readable medium" can be any apparatus that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device. The computer-readable medium can be, for example but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device. More specific examples (a non-exhaustive list) of the computer-readable medium include the following: an electrical connection (electronic) having one or more wires, a portable computer diskette (magnetic), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber (optical), and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium can even be paper or another suitable medium upon which the program is printed, as the program can be electronically captured, for example via an optical scanner, then compiled, interpreted, or otherwise processed, and stored in a computer memory in a form that can be later executed by a computer. In this context, a "computer-readable medium" can be any means that can store the program for use by or in connection with the instruction execution system, apparatus, or device.

[0152] The foregoing description of various embodiments of the application has been presented for the purposes of illustration and description. It is not intended to be exhaustive or to limit the application to the precise form disclosed, and various modifications and variations are possible in light of the above teachings. It is intended that the scope of the application be limited not with this detailed description, but rather by the claims appended hereto.

[0153] It will be appreciated that portions of the present application can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented in hardware, and in another embodiment, any of the following technologies known in the art, or a combination thereof, can be used: discrete logic circuitry having logic gates for implementing logic functions upon an application of data signals, application specific integrated circuits having appropriate combinational logic gates, programmable gate arrays (PGA), field programmable gate arrays (FPGA), and the like.

[0154] In the description of the present application, the use of the terms "one embodiment", "some embodiments", "an example", "a specific example", or "some examples" to describe certain aspects of the application, is not meant to limit or restrict the scope of the application to such embodiments, examples, or examples alone. Additionally, the description of the specific features, structures, materials, or characteristics is not meant to be an exhaustive list of all such features, structures, materials, or characteristics for the application. Rather, the description is intended to exemplify some of the many possible embodiments or examples of the application.

[0155] While the embodiments of the application have been shown and described, it is to be understood that the embodiments described are merely exemplary and are not to be taken as limiting the scope of the application. The scope of the application is defined by the claims and their equivalents.

[0156] The above is a specific description of the preferred embodiment of the present application, but the present application is not limited to the described embodiment, and those skilled in the art can make various equivalent modifications or replacements without departing from the spirit of the present application, and these equivalent modifications or replacements are all included in the scope defined by the claims of the present application.

Claims

1. A method for dynamically organizing teaching resources driven by causality under the support of teaching and learning behavior data, characterized in that, The method comprises the following steps: acquiring online teaching and learning behavior data, the online teaching and learning behavior data comprising teaching activity data, student learning data and online education resource information; modeling resource access dynamic time sequence causality according to the online teaching and learning behavior data, and determining causality information of online education resource access; predicting online education resource access heat through a neural network according to the causality information; constructing a file-level teaching and learning behavior data supported causality driven teaching resource dynamic organization process according to the predicted online education resource access heat, so as to realize causality driven teaching resource dynamic organization under the support of teaching and learning behavior data; the acquiring online teaching and learning behavior data comprises: collecting course outline and time information in a service area as teaching activity data; collecting student class information and network behavior information in the service area as student learning data; collecting online education resource attribute information and access record information as online education resource information; selecting target features from the teaching activity data, student learning data and online education resource information according to a target task scenario, and assembling observable multi-dimensional time series as source information for online education resource access feature analysis; the modeling resource access dynamic time sequence causality according to the online teaching and learning behavior data, and determining causality information of online education resource access comprises: performing boundary division on a time period with locally stable causality of resource access according to a boundary divider based on a probability distribution, to obtain online resource access dynamic causality boundary information; constructing online resource access dynamic causality description according to the online resource access dynamic causality boundary information; performing online resource access dynamic causality discovery according to the online resource access dynamic causality boundary information and the online resource access dynamic causality description, to determine causality information of online education resource access; the predicting online education resource access heat through a neural network according to the causality information comprises: acquiring a dynamic causality structure set, and constructing a causality neural network layer for online education resource access heat prediction network; constructing an online resource access heat prediction model according to the causality neural network layer; predicting online education resource access heat through the online resource access heat prediction model; the constructing an online resource access heat prediction model according to the causality neural network layer comprises: combining a time sequence message passing neural network to construct an online education resource access process representation data information propagation update network; constructing a time sequence encoding-decoding architecture based dynamic causality representation learning mechanism based on a propagation mechanism Mask method, to learn causality representation; constructing an online teaching and learning behavior state evolution network and an online resource access reaction state network according to the learned causality representation, to acquire online education resource access state; designing a time sequence differential neural network based online resource access heat predictor through the online education resource access state by using a time sequence differential network; the constructing a file-level teaching and learning behavior data supported causality driven teaching resource dynamic organization process according to the predicted online education resource access heat comprises: Obtaining data storage resource space situation, network load topology situation, and each data node system load situation in each service area; According to the predicted online education resource access heat, obtaining online education resource access situation in each service area; According to the network load topology situation and the online education resource access situation in each service area, estimating network load situation in each service of the online education system in a future period of time; According to the online education resource access situation in each service area, determining online education resource file copy quantity in each service area according to predicted access heat size; According to the data storage resource space situation in each service area, the each data node system load situation, the online education resource file copy quantity in each service area, and the network load situation in each service of the online education system in a future period of time, calculating execution resource consumption situation of the online education resource file copy adjustment system; According to the online education resource file copy quantity and the execution resource consumption situation, calculating online education resource file copy final quantity in each service area, and performing corresponding file copy generation and deletion update operation to complete online education resource organization adjustment.

2. A system for implementing the method of dynamically organizing teaching resources under the causally-driven instruction with behavior data of claim 1, wherein, Comprise: A first module for obtaining online teaching and learning behavior data, the online teaching and learning behavior data including teaching activity data, student learning data and online education resource information; A second module for modeling resource access dynamic time sequence causality according to the online teaching and learning behavior data, and determining causality information of online education resource access; A third module for predicting online education resource access heat through a neural network according to the causality information; A fourth module for constructing a file-level teaching resource dynamic organization process driven by causality under the support of teaching and learning behavior data to realize teaching resource dynamic organization driven by causality under the support of teaching and learning behavior data.

3. An electronic device, comprising: Comprise a processor and a memory; The memory is used to store a program; The processor executes the program to realize the method of claim 1.

4. A computer-readable storage medium, characterized in that, The storage medium stores a program, and the program is executed by the processor to realize the method of claim 1.

5. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to realize the method of claim 1.

Images