CN117149988B

CN117149988B - Data management processing method and system based on education digitization

Info

Publication number: CN117149988B
Application number: CN202311438571.9A
Authority: CN
Inventors: 陈志雄
Original assignee: Guangzhou Vensi Intelligent Technology Co ltd
Current assignee: Guangzhou Vensi Intelligent Technology Co ltd
Priority date: 2023-11-01
Filing date: 2023-11-01
Publication date: 2024-02-27
Anticipated expiration: 2043-11-01
Also published as: CN117149988A

Abstract

According to the data management processing method and system based on education digitization, provided by the embodiment of the invention, through rough digital education service session text semantic pairing, historical digital education service session logs are extracted, and scheme timeliness is improved. Subsequent joint analysis uses fine semantics for structured storage aid decisions. In view of the fact that semantic features carried by fine digital education service session text semantics are more complete and accurate, the structured storage auxiliary decision-making result of the digital education service session log to be managed can be accurately and reliably determined, and therefore structured storage of the digital education service session log to be managed is achieved by means of the target digital education service session log and the structured storage strategy included in the structured storage auxiliary decision-making result, structured storage accuracy and efficiency of the digital education service session log to be managed are improved, and management quality of the digital education service session log is guaranteed.

Description

Data management processing method and system based on education digitization

Technical Field

The invention relates to the technical field of data processing, in particular to a data management processing method and system based on education digitization.

Background

Through education digitization, teachers can design and teach courses by means of tools such as electronic teaching plans, multimedia resources and online learning platforms, personalized learning experience is provided, and student performance assessment is performed. Students can acquire teaching materials through the Internet, participate in online discussion, complete homework and examination, and acquire timely feedback. Meanwhile, education digitization can also provide learning opportunities of crossing regions and crossing cultures, and knowledge vision of students is widened. Education digitization also includes digitization of school management and education administration aspects, such as student information management systems, online entry and achievement management, and digitization tools for education policy formulation and supervision. Education digitization aims to improve the quality, efficiency and accessibility of education by using advanced information and communication technologies, provide wider possibilities for education and richer learning opportunities for students.

With the popularization of education digitization, data management requirements for education digitization are increasing, and efficient and accurate data storage management requirements are becoming an important problem.

Disclosure of Invention

In order to solve the technical problems in the related art, the invention provides a data management processing method and system based on education digitization.

In a first aspect, an embodiment of the present invention provides a data management processing method based on education digitization, applied to a data management processing system, where the method includes:

acquiring digital education service session text semantics of a first text attribute value and digital education service session text semantics of a second text attribute value of a digital education service session log to be managed, and digital education service session text semantics of the first text attribute value and digital education service session text semantics of the second text attribute value of each historical digital education service session log in a first digital education service session log set, wherein the first text attribute value is smaller than the second text attribute value, and the digital education service session log to be managed is a digital education service session log to be subjected to text structured storage;

carrying out joint analysis on the digital education service session text semantics of the first text attribute value of the digital education service session log to be managed and the digital education service session text semantics of the first text attribute value of each historical digital education service session log in the first digital education service session log set to obtain a corresponding first adaptation coefficient, and selecting historical digital education service session logs with target numbers from the first digital education service session log set according to the first adaptation coefficient so as to generate a second digital education service session log set;

Carrying out joint analysis on the text semantics of the digital education service session of the second text attribute value of the digital education service session log to be managed and the text semantics of the digital education service session of the second text attribute value of each historical digital education service session log in the second digital education service session log set to obtain a corresponding second adaptation coefficient, and determining a structured storage auxiliary decision result of the digital education service session log to be managed according to the second adaptation coefficient; the structured storage assistant decision result comprises a target digital education service session log which meets the adaptation condition with the digital education service session log to be managed and a structured storage policy corresponding to the target digital education service session log, wherein the target digital education service session log is one of the historical digital education service session logs in the second digital education service session log set.

In some examples, the obtaining digital education service session text semantics of the first text attribute value and the digital education service session text semantics of the second text attribute value of the digital education service session log to be managed includes:

Changing the text granularity of the to-be-managed digital education service session log to obtain a first adjusted digital education service session log and a second adjusted digital education service session log corresponding to the to-be-managed digital education service session log;

and respectively carrying out semantic mining operation on the first adjusted digital education service session log and the second adjusted digital education service session log corresponding to the digital education service session log to be managed to obtain digital education service session text semantics of the first text attribute value and digital education service session text semantics of the second text attribute value of the digital education service session log to be managed.

In some examples, the semantic mining operation is performed on the first adjusted digital education service session log and the second adjusted digital education service session log corresponding to the to-be-managed digital education service session log to obtain digital education service session text semantics of the first text attribute value and digital education service session text semantics of the second text attribute value of the to-be-managed digital education service session log, including:

respectively carrying out initial semantic mining operation on the first adjusted digital education service session log and the second adjusted digital education service session log of the digital education service session log to be managed to obtain a first initial semantic vector and a second initial semantic vector of the digital education service session log to be managed;

Performing semantic refining operation on the first initial semantic vector and the second initial semantic vector of the digital education service session log to be managed respectively to obtain a first text semantic refining vector and a second text semantic refining vector corresponding to the digital education service session log to be managed;

and respectively carrying out semantic sampling operation on the first text semantic refining vector and the second text semantic refining vector to obtain digital education service session text semantics of the first text attribute value and digital education service session text semantics of the second text attribute value corresponding to the digital education service session log to be managed.

In some examples, the semantic mining operation is implemented through a deep structured semantic network comprising an initial semantic vector mining branch, a text semantic refining branch, and a second linkage processing branch;

the semantic mining operation is performed on the first adjusted digital education service session log and the second adjusted digital education service session log corresponding to the to-be-managed digital education service session log to obtain digital education service session text semantics of the first text attribute value and digital education service session text semantics of the second text attribute value of the to-be-managed digital education service session log, including:

Performing initial semantic mining operation on the first adjusted digital education service session log and the second adjusted digital education service session log through the initial semantic vector mining branch to obtain a first initial semantic vector and a second initial semantic vector of the digital education service session log to be managed;

performing semantic refining operation on the first initial semantic vector and the second initial semantic vector through the text semantic refining branch to obtain a first text semantic refining vector and a second text semantic refining vector corresponding to the digital education service session log to be managed;

performing semantic sampling operation on the first text semantic refining vector through the second linkage processing branch to obtain digital education service session text semantics of a first text attribute value corresponding to the digital education service session log to be managed;

and carrying out semantic downsampling on the second text semantic refining vector to obtain digital education service session text semantics of a second text attribute value corresponding to the digital education service session log to be managed.

In some examples, the deep structured semantic network further comprises a first linkage processing branch, the method further comprising:

Acquiring a general deep structured semantic network and a digital education service session log example;

debugging the initial semantic vector mining branch in the general deep structured semantic network according to the digital education service session log example to obtain a first deep structured semantic network corresponding to the general deep structured semantic network;

maintaining the variable of the initial semantic vector mining branch in the first deep structured semantic network unchanged, and debugging the first linkage processing branch in the first deep structured semantic network to obtain a second deep structured semantic network corresponding to the first deep structured semantic network;

maintaining the variable of the initial semantic vector mining branch in the first deep structured semantic network and the variable of the first linkage processing branch in the first deep structured semantic network unchanged, debugging the second linkage processing branch in the second deep structured semantic network to obtain a third deep structured semantic network corresponding to the second deep structured semantic network, and taking the third deep structured semantic network as the deep structured semantic network.

In some examples, the debugging the initial semantic vector mining branch in the generic deep structured semantic network according to the digital education service session log example to obtain a first deep structured semantic network corresponding to the generic deep structured semantic network includes:

changing the text granularity of the digital education service session log example to obtain a first adjusted digital education service session log and a second adjusted digital education service session log corresponding to the digital education service session log example, wherein the first adjusted digital education service session log corresponding to the digital education service session log example is taken as an initial debugging example, the second adjusted digital education service session log corresponding to the digital education service session log example is taken as a positive example, and the rest digital education service session log examples are taken as negative examples;

performing initial semantic mining operation on the initial debugging examples, the positive examples and the negative examples through the initial semantic vector mining branches in the general deep structured semantic network to obtain corresponding initial debugging example semantics, positive example semantics and negative example semantics;

Acquiring a first semantic commonality value between the initial debugging example semantics and the positive example semantics and a second semantic commonality value between the initial debugging example semantics and the negative example semantics, and generating a first debugging cost of the general deep structured semantic network according to the first semantic commonality value and the second semantic commonality value;

and performing variable improvement on the initial semantic vector mining branch in the general deep structured semantic network according to the first debugging cost to obtain a first deep structured semantic network corresponding to the general deep structured semantic network.

In some examples, the debugging the first linkage processing branch in the first deep structured semantic network to obtain a second deep structured semantic network corresponding to the first deep structured semantic network includes:

performing semantic sampling operation on the initial debugging sample semantics, the positive sample semantics and the negative sample semantics through the first linkage processing branch in the first deep structured semantic network respectively to obtain corresponding initial debugging sample semantics, positive sample semantics and negative sample semantics;

Acquiring a third semantic commonality value between the initial debug example sampling semantic and the positive example sampling semantic and a fourth semantic commonality value between the initial debug example sampling semantic and the negative example sampling semantic, and generating a second debug cost of the first deep structured semantic network according to the third semantic commonality value and the fourth semantic commonality value;

and performing variable improvement on the first linkage processing branch in the first deep structured semantic network according to the second debugging cost to obtain a second deep structured semantic network corresponding to the first deep structured semantic network.

In some examples, the debugging the second linkage processing branch in the second deep structured semantic network to obtain a third deep structured semantic network corresponding to the second deep structured semantic network includes:

acquiring a text semantic example set, wherein the text semantic example set comprises a text semantic example corresponding to a plurality of digital education service session log examples and prior semantic annotations corresponding to the text semantic examples, and the prior semantic annotations are used for indicating sampling semantics for performing semantic sampling operation on the text semantic examples through the second linkage processing branches;

Carrying out semantic sampling operation on each text semantic example through the second linkage processing branch in the second deep structured semantic network to obtain sampling semantic prediction results corresponding to each text semantic example;

determining semantic commonality values of the sampling semantic prediction results and the corresponding priori semantic annotations respectively, and carrying out averaging operation on the semantic commonality values to obtain third debugging cost of the deep structured semantic network;

and performing variable improvement on the second linkage processing branch in the second deep structured semantic network according to the third debugging cost to obtain a third deep structured semantic network corresponding to the second deep structured semantic network.

In some examples, obtaining digital educational service session text semantics of the first text attribute value and digital educational service session text semantics of the second text attribute value for each historical digital educational service session log in the first set of digital educational service session logs comprises:

the method comprises the following steps of:

Changing the text granularity of the historical digital education service session log to obtain a third adjusted digital education service session log corresponding to the historical digital education service session log;

performing initial semantic mining operation on a third adjusted digital education service session log corresponding to the historical digital education service session log to obtain a third initial semantic vector corresponding to the historical digital education service session log;

performing semantic refining operation on a third initial semantic vector corresponding to the historical digital education service session log to obtain a third text semantic refining vector corresponding to the historical digital education service session log;

and carrying out semantic sampling operation of different levels on the third text semantic refining vector to obtain digital education service session text semantics of the first text attribute value and digital education service session text semantics of the second text attribute value corresponding to the historical digital education service session log.

In some examples, performing an initial semantic mining operation on a third adjusted digital education service session log corresponding to the historical digital education service session log to obtain a third initial semantic vector corresponding to the historical digital education service session log includes:

Performing initial semantic mining operation on a third adjusted digital education service session log corresponding to the historical digital education service session log through an AI algorithm to obtain a plurality of text semantic relation networks corresponding to the historical digital education service session log, and taking the text semantic relation networks as a third initial semantic vector;

the semantic refining operation is performed on the third initial semantic vector corresponding to the historical digital education service session log to obtain a third text semantic refining vector corresponding to the historical digital education service session log, which comprises the following steps:

and carrying out semantic aggregation operation on the text semantic relation networks to obtain a third text semantic refining vector corresponding to the historical digital education service session log.

In some examples, the performing semantic sampling operations on the third text semantic refinement vector at different levels to obtain digital education service session text semantics of the first text attribute value and digital education service session text semantics of the second text attribute value corresponding to the historical digital education service session log includes:

respectively acquiring a first sliding window scanning variable and a second sliding window scanning variable for carrying out sliding window scanning on text semantic attribute values of the third text semantic refining vector;

Carrying out semantic feature operation on the first sliding window scanning variable and the third text semantic refining vector to obtain a first semantic feature operation result, and carrying out semantic feature operation on the second sliding window scanning variable and the third text semantic refining vector to obtain a second semantic feature operation result;

and carrying out semantic feature mapping processing on the first semantic feature operation result to obtain digital education service session text semantics of a first text attribute value corresponding to the historical digital education service session log, and carrying out semantic feature mapping processing on the second semantic feature operation result to obtain digital education service session text semantics of a second text attribute value corresponding to the historical digital education service session log.

In a second aspect, the present invention also provides a data management processing system, including a processor and a memory; the processor is in communication with the memory, and the processor is configured to read and execute a computer program from the memory to implement the method described above.

In a third aspect, the present invention also provides a computer-readable storage medium having stored thereon a program which, when executed by a processor, implements the method described above.

In the process of carrying out structural storage on the digital education service session logs to be managed, carrying out joint analysis on digital education service session text semantics of first text attribute values of the digital education service session logs to be managed and digital education service session text semantics of first text attribute values of each historical digital education service session log in a first digital education service session log set to obtain corresponding first adaptation coefficients, and selecting historical digital education service session logs with target numbers from the first digital education service session log set based on the first adaptation coefficients so as to generate a second digital education service session log set; and then carrying out joint analysis on the text semantics of the digital education service session of the second text attribute value of the digital education service session log to be managed and the text semantics of the digital education service session of the second text attribute value of each historical digital education service session log in the second digital education service session log set to obtain a corresponding second adaptation coefficient, determining a structured storage auxiliary decision result of the digital education service session log to be managed based on the second adaptation coefficient, wherein the structured storage auxiliary decision result comprises a target digital education service session log meeting adaptation conditions with the digital education service session log to be managed and a structured storage strategy corresponding to the target digital education service session log, and the target digital education service session log is one of the historical digital education service session logs in the second digital education service session log set. In view of the fact that the first text attribute value is smaller than the second text attribute value, a certain number of historical digital education service session logs can be extracted through rough digital education service session text semantic pairing during primary joint analysis, and timeliness of the whole scheme can be improved. Further, in the subsequent joint analysis process, for a certain number of extracted historical digital education service session logs, a structural storage auxiliary decision result can be determined based on fine digital education service session text semantics through joint analysis, and in view of the fact that semantic features carried by the fine digital education service session text semantics are more complete and accurate, the structural storage auxiliary decision result of the digital education service session logs to be managed can be accurately and reliably determined, so that structural storage of the digital education service session logs to be managed is realized by utilizing target digital education service session logs and structural storage strategies included in the structural storage auxiliary decision result, structural storage accuracy and efficiency of the digital education service session logs to be managed are improved, and management quality of the digital education service session logs is guaranteed.

Drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and together with the description, serve to explain the principles of the invention.

Fig. 1 is a schematic flow chart of a data management processing method based on education digitization according to an embodiment of the present invention.

Description of the embodiments

Reference will now be made in detail to exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, the same numbers in different drawings refer to the same or similar elements, unless otherwise indicated. The implementations described in the following exemplary examples do not represent all implementations consistent with the invention. Rather, they are merely examples of apparatus and methods consistent with aspects of the invention.

It should be noted that the terms "first," "second," and the like in the description of the present invention and the above-described drawings are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order.

The method embodiments provided by the embodiments of the present invention may be performed in a data management processing system, a computer device, or a similar computing device. Taking the example of running on a data management processing system, the data management processing system may comprise one or more processors (which may include, but is not limited to, a processing means such as a microprocessor MCU or a programmable logic device FPGA) and a memory for storing data, and optionally the data management processing system may further comprise transmission means for communication functions. It will be appreciated by those of ordinary skill in the art that the above-described architecture is merely illustrative and is not intended to limit the architecture of the data management processing system described above. For example, the data management processing system may also include more or fewer components than shown above, or have a different configuration than shown above.

The memory may be used to store a computer program, for example, a software program of application software and a module, such as a computer program corresponding to a data management processing method based on education digitization in the embodiment of the present invention, and the processor executes various functional applications and data processing by running the computer program stored in the memory, that is, implements the above-mentioned method. The memory may include high speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid state memory. In some examples, the memory may further include memory remotely located with respect to the processor, the remote memory being connectable to the data management processing system through a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.

The transmission means is used for receiving or transmitting data via a network. Specific examples of the network described above may include a wireless network provided by a communication provider of a data management processing system. In one example, the transmission means comprises a network adapter (Network Interface Controller, simply referred to as NIC) that can be connected to other network devices via a base station to communicate with the internet. In one example, the transmission device may be a Radio Frequency (RF) module, which is used to communicate with the internet wirelessly.

Based on this, referring to fig. 1, fig. 1 is a flow chart of a data management processing method based on education digitization according to an embodiment of the present invention, where the method is applied to a data management processing system, and further may include S101-S103.

S101, acquiring digital education service session text semantics of a first text attribute value and digital education service session text semantics of a second text attribute value of a digital education service session log to be managed, and digital education service session text semantics of the first text attribute value and digital education service session text semantics of the second text attribute value of each historical digital education service session log in a first digital education service session log set.

The text attribute value can be understood as a semantic feature dimension, the first text attribute value is smaller than the second text attribute value, and the digital education service session log to be managed is a digital education service session log to be subjected to text structured storage. Further, digital education service session text semantics of a first text attribute value of the digital education service session log to be managed are used for characterizing coarser text features of the digital education service session log to be managed, and digital education service session text semantics of a second text attribute value of the digital education service session log to be managed are used for characterizing finer text features of the digital education service session log to be managed. Similarly, the digital education service session text semantics of the first text attribute value of each historical digital education service session log are used to characterize the coarser text features of each historical digital education service session log, and the digital education service session text semantics of the second text attribute value of each historical digital education service session log are used to characterize the finer text features of each historical digital education service session log.

In the embodiment of the invention, the digital education service session log refers to an operation interaction text record between the remote education client and the remote education server. Further, one example of a digital educational service session log is as follows:

timestamp: 2023-09-0615:30:00;

user ID:123456;

operation type: submitting a problem;

the content is as follows: i do not understand how this topic solves?

In this example, the digital educational service session log records the operations performed by a particular user (ID 123456) at a specified time (3:30 PM on 9/2023/6/2023) that a question was submitted, the specific content of which was confusing about the relevant topic.

In further examples, the digital educational service session log is a text record that records the digital educational service program during operation. Based on this, the following are some example information that may be contained in the digital educational service session log:

1) User information: recording identity information, user names, roles and the like of users;

2) Operation record: including various operations performed by the user in the digital education service, such as browsing courses, submitting jobs, asking questions, etc.;

3) Interaction information: recording interaction information between the user and the system or other users, such as chat records, discussion area contents and the like;

4) Learning progress: recording progress information of a user in a learning process, such as the number of completed courses, learning duration and the like;

5) Errors and anomalies: recording errors, abnormal conditions and corresponding error codes or description information of the system;

6) Log timestamp: the generation time of each log is recorded and used for tracking the occurrence sequence and the time line of the event.

For another example, the digital educational service session log to be managed may include the following: user A (ID: 123456) logged into the digital education service platform at 10 A.M. at 2023, 9, 7; the user A browses the content of the course mathematics foundation and stays for 20 minutes on the video learning page; user A submits post-class operation with HW001 number; user a has published a question in the discussion area: "how to ask how to solve the problem? "user B replied to user A's problem and provided a solution; the system records that the progress of the user A in completing the course of mathematics foundation is 30%; when user A submits a job, the system makes an error (error code: 500), and records error information: "unable to upload files, please try again later"; log timestamp: 2023, 9, 7, 10 am, 15 minutes. The above examples illustrate various information that may be contained in a digital educational service session log for recording a user's activities, system operation, interactions with other users, and the like.

Based on the above example, the digital education service session text semantics of the first text attribute value of the digital education service session log to be managed may be relatively brief semantic features for implementing preliminary screening of the digital education service session log, and the digital education service session text semantics of the second text attribute value of the digital education service session log to be managed may be relatively complex semantic features for implementing fine screening of the digital education service session log. The text semantics of the digital education service session in the embodiment of the invention can be represented by feature vectors (such as integer type features or floating point type features), and specific feature vector values can be flexibly selected according to actual situations, which is not limited herein.

S102, carrying out joint analysis on the digital education service session text semantics of the first text attribute value of the digital education service session log to be managed and the digital education service session text semantics of the first text attribute value of each historical digital education service session log in the first digital education service session log set to obtain a corresponding first adaptation coefficient, and selecting the historical digital education service session logs of the target number from the first digital education service session log set according to the first adaptation coefficient so as to generate a second digital education service session log set.

The joint analysis may be understood as a matching analysis, and the adaptation coefficient may be understood as a matching degree. Further, the joint analysis of the text semantics of the digital education service session based on the first text attribute value can be understood as a preliminary screening process, so that each historical digital education service session log in the first digital education service session log set is rapidly scanned, and the historical digital education service session logs with the target number are selected through the first adaptation coefficient, so that a second digital education service session log set with the relatively small number is obtained.

S103, carrying out joint analysis on the digital education service session text semantics of the second text attribute value of the digital education service session log to be managed and the digital education service session text semantics of the second text attribute value of each historical digital education service session log in the second digital education service session log set to obtain a corresponding second adaptation coefficient, and determining a structured storage auxiliary decision result of the digital education service session log to be managed according to the second adaptation coefficient.

The structured storage assistant decision result comprises a target digital education service session log which meets the adaptation condition with the digital education service session log to be managed and a structured storage policy corresponding to the target digital education service session log, wherein the target digital education service session log is one of the historical digital education service session logs in the second digital education service session log set.

It will be appreciated that after the second digital educational service session log set is obtained, a fine screening process may be implemented in accordance with digital educational service session text semantics of the second text attribute value, thereby determining a structured storage aid decision result based on the second adaptation coefficient. In the embodiment of the invention, the adaptation coefficient can also be expressed by cosine similarity.

In other words, based on the technical scheme, the following beneficial effects can be achieved:

(1) The screening efficiency is improved: by performing a joint analysis of digital educational service session text semantics of the first text attribute value of the digital educational service session log to be managed, a second set of digital educational service session logs can be quickly initially screened out. Therefore, the number of logs needing to be further analyzed and processed can be reduced, and time and resources are saved;

(2) And (3) improving screening precision: on the basis of the preliminary screening, the digital education service session logs meeting specific conditions or requirements can be further screened out by carrying out joint analysis on the digital education service session text semantics of the second text attribute value of the digital education service session log to be managed and the digital education service session text semantics of the second text attribute value of each historical digital education service session log in the second digital education service session log set. Thus, the selected logs can be ensured to be more accurate and relevant;

(3) Generating a structured storage aid decision result: and determining a structured storage aid decision result of the digital education service session log to be managed according to the first adaptation coefficient and the second adaptation coefficient. The decision result includes a target digital educational service session log and corresponding structured storage policy that satisfy the fit condition. This may provide a more specific and practical decision scheme for managing digital educational service session logs, helping to organize and utilize the value of these logs;

(4) Optimizing resource utilization: unnecessary analysis and processing can be avoided through the process of primary screening and fine screening, reducing the need for computing resources and storage space. Selecting only the target digital educational service session log for further processing can more efficiently utilize limited resources and reduce associated costs.

In some alternative embodiments, the digital education service session text semantics of acquiring the first text attribute value and the second text attribute value of the log of digital education service sessions to be managed in S101 includes S1011 and S1012.

S1011, changing the text granularity of the digital education service session log to be managed to obtain a first adjusted digital education service session log and a second adjusted digital education service session log corresponding to the digital education service session log to be managed.

Wherein the text fine granularity is used for representing the content richness of the digital education service session log to be managed. The higher the text granularity, the richer the content of the digital education service session log to be managed, the lower the text granularity, and the more brief the content of the digital education service session log to be managed. Further, a first adjusted digital education service session log corresponds to the digital education service session log of the first text attribute value and a second adjusted digital education service session log corresponds to the digital education service session log of the second text attribute value.

Taking the digital education service session log to be managed as a section of content for describing student questions and feedback in detail as an example, the log is subjected to text fine-grained modification, and is adjusted to a first adjusted digital education service session log and a second adjusted digital education service session log. The first adjusted digital education service session log may be a short description of the summary process that extracts the major questions and key information. While a second adjusted digital educational service session log may be filtered and categorized to retain only content related to a particular topic or domain.

S1012, semantic mining operations are respectively carried out on the first adjusted digital education service session log and the second adjusted digital education service session log corresponding to the digital education service session log to be managed, so that digital education service session text semantics of the first text attribute value and digital education service session text semantics of the second text attribute value of the digital education service session log to be managed are obtained.

For the first adjusted digital education service session log, the semantics thereof may be mined using natural language processing techniques such as text classification, keyword extraction, or emotion analysis. Through these techniques, the topics, important questions, or keywords to which the journal pertains can be known and the text semantics of the digital educational service session associated therewith can be obtained. Likewise, similar semantic mining techniques may be applied to the second adjusted digital education service session log. These techniques may help understand the semantic meaning of a particular domain or topic covered by the log, further mining information and insight therein.

In summary, through S1011 and S1012, fine-grained adjustment is performed on the digital education service session logs to be managed, and the digital education service session text semantics of the first text attribute value and the second text attribute value are obtained through a semantic mining operation, which helps to better understand and utilize the information in these logs.

In addition, at least the following advantageous effects can be achieved through S1011 and S1012:

(1) Personalized adjustment and excavation: through text fine granularity adjustment of the digital education service session log to be managed, the log content can be subjected to personalized adjustment according to specific requirements and targets, and a first adjusted digital education service session log and a second adjusted digital education service session log are generated. Therefore, specific requirements and application scenes can be better met, and log data with more pertinence and practicability are provided;

(2) Simplifying information processing: by performing a semantic mining operation on the first adjusted digital education service session log and the second adjusted digital education service session log, key information, topics, and semantic meanings may be extracted from the log. Thus, the processing and analysis work of a large number of original logs is reduced, the needed information is more clear and easy to understand, and the complexity of information processing is simplified;

(3) The excavation efficiency and accuracy are improved: through semantic mining operations, natural language processing techniques and machine learning algorithms can be employed to understand the semantics of log text and hidden information in depth. Thus, valuable knowledge and hole finding can be rapidly and automatically extracted from a large amount of data, the mining efficiency is improved, and more accurate and reliable semantic analysis results are provided;

(4) Customized applications and decision support: and obtaining the text semantics of the digital education service session of the first text attribute value and the second text attribute value through personalized adjustment and semantic mining. This may provide customized data support for a particular application scenario, problem solution, or decision. By deeply understanding the meaning and context of log information, application development, business decision and innovation research in the related fields can be better supported.

In summary, the personalized adjustment and semantic mining operations in S1011 and S1012 bring about various beneficial effects, including improving information processing efficiency, accuracy and custom application capability, and providing a more flexible and intelligent solution for the management and analysis of digital education service session logs.

In some examples, semantic mining operations are performed on the first adjusted digital education service session log and the second adjusted digital education service session log corresponding to the to-be-managed digital education service session log in S1012 to obtain digital education service session text semantics of the first text attribute value and digital education service session text semantics of the second text attribute value of the to-be-managed digital education service session log, including S10121-S10123.

S10121, respectively performing initial semantic mining operation on the first adjusted digital education service session log and the second adjusted digital education service session log of the to-be-managed digital education service session log to obtain a first initial semantic vector and a second initial semantic vector of the to-be-managed digital education service session log.

S10122, respectively carrying out semantic refining operation on the first initial semantic vector and the second initial semantic vector of the digital education service session log to be managed to obtain a first text semantic refining vector and a second text semantic refining vector corresponding to the digital education service session log to be managed.

The semantic refining operation is a pooling operation, and the text semantic refining vector is a semantic pooling vector.

S10123, respectively carrying out semantic sampling operation on the first text semantic refining vector and the second text semantic refining vector to obtain digital education service session text semantics of the first text attribute value and digital education service session text semantics of the second text attribute value corresponding to the digital education service session log to be managed.

The semantic sampling operation is semantic sampling processing.

In S1011, a first adjusted digital education service session log and a second adjusted digital education service session log have been obtained. Now, according to S10121, an initial semantic mining operation is performed on the two adjusted logs, to obtain a first initial semantic vector and a second initial semantic vector of the digital education service session log to be managed.

It is assumed that the first adjusted digital education service session log is a piece of text describing a student problem and solution, and the second adjusted digital education service session log contains only information related to a specific topic. Through initial semantic mining operations, the two logs are subjected to semantic analysis by using natural language processing technologies such as word embedding, topic modeling or emotion analysis. Thus, a first initial semantic vector and a second initial semantic vector may be generated for representing initial semantic information of the log text.

According to S10122, the semantic refining operation is further performed on the first initial semantic vector and the second initial semantic vector to obtain a first text semantic refining vector and a second text semantic refining vector of the digital education service session log to be managed. The semantic refining operation employs a pooling operation, i.e., aggregating vector data into a single value.

For the first initial semantic vector and the second initial semantic vector, a pooling method, such as average pooling or maximum pooling, may be applied to aggregate the information in the vectors. The first text semantic refining vector and the second text semantic refining vector thus obtained represent the overall semantic meaning of the respective logs, but are represented by a single value.

In S10123, according to S10123, performing a semantic sampling operation on the first text semantic refining vector and the second text semantic refining vector to obtain digital education service session text semantics of the first text attribute value and digital education service session text semantics of the second text attribute value of the digital education service session log to be managed. The semantic sampling operation is to sample the semantic vector.

It is assumed that the first text semantic refining vector represents the importance of the student question and the second text semantic refining vector represents the depth of the subject in question. Specific semantic information can be obtained from the two vectors through semantic sampling operations. For example, it is possible to prioritize student questions by comparing the size of the first text semantic refining vector. Likewise, the degree of relevance of the discussion topic may be determined by comparing the values of the second text semantic refining vector.

In summary, the operations in S10121 to S10123 perform multi-level semantic processing on the digital education service session log to be managed through initial semantic mining, semantic refining and semantic sampling. These operations facilitate extracting semantic information of the journal and obtaining therefrom text semantics of the digital educational service session for the particular text attribute value. Such a process flow may provide a more specific, accurate, and meaningful semantic representation for further analysis and application.

In an embodiment of the present invention, the semantic mining operation is implemented through a deep structured semantic network comprising an initial semantic vector mining branch, a text semantic refining branch, and a second linkage processing branch (second matching branch). Based on this, performing semantic mining operation on the first adjusted digital education service session log and the second adjusted digital education service session log corresponding to the to-be-managed digital education service session log in S1012 to obtain digital education service session text semantics of the first text attribute value and digital education service session text semantics of the second text attribute value of the to-be-managed digital education service session log, including: performing initial semantic mining operation on the first adjusted digital education service session log and the second adjusted digital education service session log through the initial semantic vector mining branch to obtain a first initial semantic vector and a second initial semantic vector of the digital education service session log to be managed; performing semantic refining operation on the first initial semantic vector and the second initial semantic vector through the text semantic refining branch to obtain a first text semantic refining vector and a second text semantic refining vector corresponding to the digital education service session log to be managed; performing semantic sampling operation on the first text semantic refining vector through the second linkage processing branch to obtain digital education service session text semantics of a first text attribute value corresponding to the digital education service session log to be managed; and carrying out semantic downsampling on the second text semantic refining vector to obtain digital education service session text semantics of a second text attribute value corresponding to the digital education service session log to be managed.

In some examples, the deep structured semantic network further comprises a first linkage processing branch. Based on this, the method further comprises S201-S204.

S201, a general deep structured semantic network and a digital education service session log example are obtained.

The universal deep structured semantic network is a deep structured semantic network to be trained. Digital educational service session log examples digital educational service session log samples.

S202, debugging the initial semantic vector mining branch in the general deep structured semantic network according to the digital education service session log example to obtain a first deep structured semantic network corresponding to the general deep structured semantic network.

S203, maintaining the variable of the initial semantic vector mining branch in the first deep structured semantic network unchanged, and debugging the first linkage processing branch in the first deep structured semantic network to obtain a second deep structured semantic network corresponding to the first deep structured semantic network.

S204, maintaining variables of the initial semantic vector mining branch in the first deep structured semantic network and variables of the first linkage processing branch in the first deep structured semantic network unchanged, debugging the second linkage processing branch in the second deep structured semantic network to obtain a third deep structured semantic network corresponding to the second deep structured semantic network, and taking the third deep structured semantic network as the deep structured semantic network.

It can be seen that by debugging the digital educational service session log example, initial semantic representations of text data can be extracted using initial semantic vector mining branches in a generic deep structured semantic network. This may increase the understanding of the model into the text so that the model can better capture semantic information of the text. In the first deep structured semantic network, the variable of the initial semantic vector mining branch is kept unchanged, and the first linkage processing branch is debugged. The effect of this is to further optimize the initial semantic vector representation, improving the semantic expression capabilities of the text by combining context information and semantic relevance. This enables the model to more accurately understand the meaning of the text and reflect it in the semantic vector. And continuously keeping the variables of the initial semantic vector mining branch and the first linkage processing branch in the first deep structured semantic network unchanged, and debugging the second linkage processing branch. The effect of this step is to further optimize the semantic vector representation, improving the semantic expressive power of the text through more advanced semantic association processing. Because the second linkage processing branch is established on the basis of the first two branches, the semantic information processed previously can be better utilized, and the understanding capability of the model on the text semantics is further improved.

In summary, through multiple debugging and optimization, a deep structured semantic network is obtained through an initial semantic vector mining branch, a first linkage processing branch and a second linkage processing branch. This network is able to more accurately extract semantic information from the input text and represent it in the form of semantic vectors. Such deep structured semantic networks have important application potential in digital educational services, such as structured storage, automated question-answering systems, learning content recommendations, etc.

In some examples, according to the digital education service session log example in S202, the initial semantic vector mining branch in the generic deep structured semantic network is debugged to obtain a first deep structured semantic network corresponding to the generic deep structured semantic network, including S2021-S2024.

S2021, changing the text granularity of the digital education service session log example to obtain a first adjusted digital education service session log and a second adjusted digital education service session log corresponding to the digital education service session log example, wherein the first adjusted digital education service session log corresponding to the digital education service session log example is used as an initial debugging example, the second adjusted digital education service session log corresponding to the digital education service session log example is used as a positive example, and the rest digital education service session log examples are used as negative examples.

S2022, performing initial semantic mining operation on the initial debugging example, the positive example and the negative example through the initial semantic vector mining branch in the general deep structured semantic network to obtain corresponding initial debugging example semantics, positive example semantics and negative example semantics.

S2023, acquiring a first semantic commonality value between the initial debugging example semantics and the positive example semantics and a second semantic commonality value between the initial debugging example semantics and the negative example semantics, and generating a first debugging cost of the general deep structured semantic network according to the first semantic commonality value and the second semantic commonality value.

S2024, performing variable improvement on the initial semantic vector mining branch in the general deep structured semantic network according to the first debugging cost to obtain a first deep structured semantic network corresponding to the general deep structured semantic network.

It will be appreciated that the application S2021-S2024 creates a first adjusted digital education service session log and a second adjusted digital education service session log by making text fine-grained changes to the digital education service session log example. The method has the advantages that session logs with different versions can be owned, more semantic expression modes and scenes are covered, and more rich training data is provided for the model. The method comprises the steps of mining branches by using initial semantic vectors in a general deep structured semantic network, and performing initial semantic mining operation on initial debugging examples, positive examples and negative examples. This will enable models to learn their semantic information from different types of examples, further enhancing the model's understanding of the text. A first semantic commonality value between the initial debug instance semantics and the positive instance semantics and a second semantic commonality value between the initial debug instance semantics and the negative instance semantics are calculated. The commonality values reflect the semantic similarity degree between different examples, can help evaluate the performance of the model when processing different types of examples, and provide guidance for subsequent tuning. And carrying out variable improvement on the initial semantic vector mining branch in the general deep structured semantic network according to the calculated first debugging cost. These improvements will enable the model to better capture the initial semantic representation of the text, enhance the performance of the model at the initial semantic mining stage, and provide more accurate and targeted semantic information for subsequent linked processing branches.

The combination of the above steps can increase the semantic understanding and processing power of the model in digital educational services. By using session log examples of different versions, rich semantic expression modes and scenes are introduced, and the model can be better adapted to various actual conditions. Meanwhile, by optimizing the initial semantic vector mining branch and generating the first debugging cost, the model can extract and represent the semantic information of the text more accurately, and a more reliable basis is provided for the subsequent processing steps.

In some optional examples, debugging the first linkage processing branch in the first deep structured semantic network in S203 obtains a second deep structured semantic network corresponding to the first deep structured semantic network, including S2031-S2033.

S2031, performing semantic sampling operation on the initial debugging sample semantics, the positive sample semantics and the negative sample semantics through the first linkage processing branch in the first deep structured semantic network to obtain corresponding initial debugging sample semantics, positive sample semantics and negative sample semantics.

S2032, obtaining a third semantic commonality value between the initial debug sample semantics and the positive example sample semantics and a fourth semantic commonality value between the initial debug sample semantics and the negative example sample semantics, and generating a second debug cost of the first deep structured semantic network according to the third semantic commonality value and the fourth semantic commonality value.

S2033, performing variable improvement on the first linkage processing branch in the first deep structured semantic network according to the second debugging cost to obtain a second deep structured semantic network corresponding to the first deep structured semantic network.

In the embodiment of the invention, the first linkage processing branch in the first deep structured semantic network is used for carrying out semantic sampling operation on the initial debugging example semantics, the positive example semantics and the negative example semantics. This will enable the model to extract key semantic information from the examples and generate corresponding sampled semantic representations. A third semantic commonality value between the initial debug sample semantics and the positive sample semantics and a fourth semantic commonality value between the initial debug sample semantics and the negative sample semantics are calculated. The commonality values reflect the similarity degree between sampling semantics, can help evaluate the performance of the model when processing different types of sampling semantics, and provide guidance for subsequent tuning. And performing variable improvement on the first linkage processing branch in the first deep structured semantic network based on the calculated second debugging cost. By optimizing parameters and weights of the linkage processing branches, the model can more accurately capture the association relation between sampling semantics, and the performance of the model in the semantic reasoning and linkage processing stages is improved.

In summary, the application of S2031-S2033 can enhance the semantic understanding and linkage processing capabilities of the model in digital educational services. Through semantic sampling operation and calculation of semantic commonality values, the model can better understand semantic similarity and difference among examples, and tuning and improvement of the model are carried out according to the information. The optimized second deep structured semantic network has higher expression capability and reasoning capability.

In some examples, debugging the second linkage processing branch in the second deep structured semantic network in S204 to obtain a third deep structured semantic network corresponding to the second deep structured semantic network includes: obtaining a text semantic example set, wherein the text semantic example set comprises a text semantic example corresponding to a plurality of digital education service session log examples and a priori semantic annotation (feature tag) corresponding to each text semantic example, and the priori semantic annotation is used for indicating sampling semantics of performing semantic sampling operation on the text semantic example through the second linkage processing branch; carrying out semantic sampling operation on each text semantic example through the second linkage processing branch in the second deep structured semantic network to obtain sampling semantic prediction results corresponding to each text semantic example; determining semantic commonality values of the sampling semantic prediction results and the corresponding priori semantic annotations respectively, and carrying out averaging operation on the semantic commonality values to obtain third debugging cost of the deep structured semantic network; and performing variable improvement on the second linkage processing branch in the second deep structured semantic network according to the third debugging cost to obtain a third deep structured semantic network corresponding to the second deep structured semantic network.

Based on the above, the performance of the third deep structured semantic network can be further improved. By using a set of textual semantic instances and a priori semantic annotations, the model may more accurately perform semantic sampling operations and generate sampled semantic predictions that are related to the a priori annotations. The calculated semantic commonality values help evaluate the model's performance in processing different examples and guide the model's improvement. The third deep structured semantic network can better understand and express text semantics through variable improvement.

In some examples, the digital education service session text semantics of acquiring the first text attribute value and the second text attribute value of each of the historical digital education service session logs in the first digital education service session log set in S101 includes: the method comprises the following steps of: changing the text granularity of the historical digital education service session log to obtain a third adjusted digital education service session log corresponding to the historical digital education service session log; performing initial semantic mining operation on a third adjusted digital education service session log corresponding to the historical digital education service session log to obtain a third initial semantic vector corresponding to the historical digital education service session log; performing semantic refining operation on a third initial semantic vector corresponding to the historical digital education service session log to obtain a third text semantic refining vector corresponding to the historical digital education service session log; and performing semantic sampling operations of different levels (different degrees) on the third text semantic refining vector to obtain digital education service session text semantics of the first text attribute value and digital education service session text semantics of the second text attribute value corresponding to the historical digital education service session log.

In acquiring text attribute values of the historical digital education service session logs in the first digital education service session log set, a series of steps are taken to obtain text semantics of the digital education service session. The steps include: text fine granularity modification: the text of each historical digital education service session log is finely adjusted to obtain a corresponding third adjusted digital education service session log. Doing so can ensure that the text better meets the semantic understanding requirements of the model; initial semantic mining: and performing initial semantic mining operation on the third adjusted digital education service session log to obtain a corresponding third initial semantic vector. This step aims at extracting key initial semantic information from the text; semantic refining: and carrying out semantic refining operation on the third initial semantic vector to obtain a third text semantic refining vector. Through further processing and optimization, the quality and the expression capacity of the text semantic vector can be improved; semantic sampling: and carrying out semantic sampling operations of different levels or degrees on the third text semantic refining vector to obtain the text semantics of the digital education service session of the first text attribute value and the text semantics of the digital education service session of the second text attribute value. Semantic information associated with different attribute values can be obtained to help the model better understand and process different types of text.

Therefore, the quality and expression capacity of text semantics can be improved by performing operations such as fine adjustment, initial semantic mining, semantic refining, semantic sampling and the like on the historical digital education service session log. This helps the model more accurately understand and process text information in the digital educational service session and extract semantic information from it that is related to different attribute values.

In other possible embodiments, the performing an initial semantic mining operation on the third adjusted digital education service session log corresponding to the historical digital education service session log to obtain a third initial semantic vector corresponding to the historical digital education service session log includes: performing initial semantic mining operation on a third adjusted digital education service session log corresponding to the historical digital education service session log through an AI algorithm to obtain a plurality of text semantic relation networks (text semantic feature graphs) corresponding to the historical digital education service session log, and taking the text semantic relation networks as a third initial semantic vector; the semantic refining operation is performed on the third initial semantic vector corresponding to the historical digital education service session log to obtain a third text semantic refining vector corresponding to the historical digital education service session log, which comprises the following steps: and carrying out semantic aggregation operation on the text semantic relation networks to obtain a third text semantic refining vector corresponding to the historical digital education service session log.

The AI algorithm can be a convolutional neural network model, and the initial semantic mining operation is performed through the AI algorithm, so that semantic information in a historical digital education service session log can be more accurately captured. And a plurality of text semantic relation networks are used as a third initial semantic vector, so that richer and comprehensive semantic association information can be contained, and the understanding capacity and the expression capacity of the model to text semantics are improved. And carrying out semantic aggregation operation on the plurality of text semantic relation networks to obtain a third text semantic refining vector. The method can reduce the dependence on a large-scale semantic graph or a complex calculation model, thereby reducing the expenditure of operation resources. Meanwhile, the semantic aggregation operation can also improve the calculation efficiency, so that semantic refining for processing a large number of historical digital education service session logs is more efficient.

In conclusion, by adopting an AI algorithm to perform initial semantic mining and semantic refining operations, the operation precision can be improved and the resource cost can be reduced. By the method, the text semantic understanding capability of the model in the historical digital education service session log can be enhanced, more accurate and comprehensive text semantic feature representation can be provided while the calculation efficiency is maintained, and more accurate and personalized support and suggestion can be provided for the subsequent digital education service.

In some examples, the performing semantic sampling operations on the third text semantic refinement vector at different levels to obtain digital education service session text semantics of the first text attribute value and digital education service session text semantics of the second text attribute value corresponding to the historical digital education service session log includes: respectively acquiring a first sliding window scanning variable and a second sliding window scanning variable for carrying out sliding window scanning on text semantic attribute values of the third text semantic refining vector; carrying out semantic feature operation on the first sliding window scanning variable and the third text semantic refining vector to obtain a first semantic feature operation result, and carrying out semantic feature operation on the second sliding window scanning variable and the third text semantic refining vector to obtain a second semantic feature operation result; and carrying out semantic feature mapping processing on the first semantic feature operation result to obtain digital education service session text semantics of a first text attribute value corresponding to the historical digital education service session log, and carrying out semantic feature mapping processing on the second semantic feature operation result to obtain digital education service session text semantics of a second text attribute value corresponding to the historical digital education service session log. The sliding window scanning comprises compression processing, the semantic feature operation comprises feature multiplication processing, and the semantic feature mapping processing comprises nonlinear processing.

The digital education service session text semantics associated with the first text attribute value and the second text attribute value may be obtained by performing a semantic sampling operation on different levels of the third text semantic refinement vector. The operations can make full use of sliding window scanning variable, semantic feature operation, semantic feature mapping processing and other means, so that the model can better express semantic information of different text attribute values. By using a specific semantic feature operation and semantic feature mapping processing method, finer and accurate semantic association can be captured, so that modeling and prediction effects on text attribute values in a historical digital education service session log are improved. The complex relationship can be better processed by adopting a nonlinear processing method, and the fitting capacity of the model is enhanced. The operation improves the semantic expression capability and simultaneously needs to reasonably consider the cost of computing resources. By compressing the sliding window scanning and adopting a nonlinear processing method, the computational complexity can be reduced and the efficiency can be improved.

Therefore, by means of semantic sampling operation of different levels of the third text semantic refining vector and combining the technologies of sliding window scanning, semantic feature operation, semantic feature mapping processing and the like, semantic modeling and prediction accuracy of text attribute values in the historical digital education service session log can be improved, and resource expenditure is controlled. This will help provide more accurate, personalized digital educational service support and advice.

The following is an example of a structured store introduction to a digital educational service session log to be managed. Assuming a digital educational service platform, each student generates a conversation log when using the platform, and records the information of questions, answers, learning progress and the like of the students. To more effectively manage these session logs, the following structured storage policies may be employed:

and (3) storing a database: the session log is stored in the database in the form of structured data. The session log may be stored using a relational database (e.g., mySQL) or a non-relational database (e.g., mongo db). By defining the proper tables and fields, the session logs can be organized and retrieved according to dimensions of students, courses, time and the like;

text analysis and classification: and carrying out text analysis and classification on the session log, extracting key information, and storing the key information in corresponding fields or labels. For example, natural language processing technology can be used for processing the problems such as word segmentation, named entity recognition and the like, and the results are stored in corresponding fields, so that the subsequent data analysis and decision support are facilitated;

association data: the session log is associated with other relevant data. For example, a student's conversation log may be associated with his personal information, learning results, etc., in order to more fully understand the student's learning situation and needs. Such an association can provide more rich background information for decisions;

Data indexing and searching: an index is built for the session log to support fast retrieval and search functions. Through reasonable index design, the data access speed can be increased, and the query efficiency of specific conditions can be improved;

data backup and security: and proper data backup measures are adopted to ensure the safety and reliability of the session log. Data is backed up regularly, and access rights and encryption measures are set at the same time so as to protect confidentiality of sensitive information;

the structured storage strategy can better manage and utilize the historical digital education service session log, thereby providing accurate and timely data support for decision making and laying a solid foundation for subsequent tasks such as data analysis, model training and the like.

The structured storage policy is a structured storage policy corresponding to the target digital education service session log, and in view of that the target digital education service session log and the digital education service session log to be managed meet the adaptation condition, the structured storage policy can be transferred to the digital education service session log to be managed.

Further, there is also provided a computer-readable storage medium having stored thereon a program which, when executed by a processor, implements the above-described method.

In the embodiments provided in the present invention, it should be understood that the disclosed apparatus and method may be implemented in other manners. The apparatus and method embodiments described above are merely illustrative, for example, flow diagrams and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of apparatus, methods and computer program products according to various embodiments of the present invention. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.

In addition, functional modules in the embodiments of the present invention may be integrated together to form a single part, or each module may exist alone, or two or more modules may be integrated to form a single part.

The functions, if implemented in the form of software functional modules and sold or used as a stand-alone product, may be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution 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 network device, or the like) 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 random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes. It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.

The above description is only of the preferred embodiments of the present invention and is not intended to limit the present invention, but various modifications and variations can be made to the present invention by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A data management processing method based on education digitization, applied to a data management processing system, the method comprising:

2. The method of claim 1, wherein the obtaining digital education service session text semantics of the first text attribute value and digital education service session text semantics of the second text attribute value of the digital education service session log to be managed comprises:

3. The method of claim 2, wherein performing semantic mining operations on the first adjusted digital education service session log and the second adjusted digital education service session log corresponding to the digital education service session log to be managed, respectively, to obtain digital education service session text semantics of the first text attribute value and digital education service session text semantics of the second text attribute value of the digital education service session log to be managed, comprises:

4. The method of claim 2, wherein the semantic mining operation is implemented through a deep structured semantic network comprising an initial semantic vector mining branch, a text semantic refining branch, and a second linkage processing branch;

5. The method of claim 4, wherein the deep structured semantic network further comprises a first linked processing branch, the method further comprising:

acquiring a general deep structured semantic network and a digital education service session log example; the universal deep structured semantic network is a deep structured semantic network to be trained;

Maintaining the variable of the initial semantic vector mining branch in the first deep structured semantic network and the variable of the first linkage processing branch in the first deep structured semantic network unchanged, debugging the second linkage processing branch in the second deep structured semantic network to obtain a third deep structured semantic network corresponding to the second deep structured semantic network, and taking the third deep structured semantic network as the deep structured semantic network;

the debugging the initial semantic vector mining branch in the general deep structured semantic network according to the digital education service session log example to obtain a first deep structured semantic network corresponding to the general deep structured semantic network, including:

6. The method of claim 5, wherein the debugging the first coordinated processing branch in the first deep-structured semantic network to obtain a second deep-structured semantic network corresponding to the first deep-structured semantic network comprises:

7. The method of claim 5, wherein the debugging the second linkage processing branch in the second deep structured semantic network to obtain a third deep structured semantic network corresponding to the second deep structured semantic network comprises:

8. The method of claim 1, wherein obtaining digital educational service session text semantics of the first text attribute value and digital educational service session text semantics of the second text attribute value for each historical digital educational service session log in the first set of digital educational service session logs comprises:

The method comprises the following steps of:

performing semantic sampling operations of different levels on the third text semantic refining vector to obtain digital education service session text semantics of a first text attribute value and digital education service session text semantics of a second text attribute value corresponding to the historical digital education service session log;

the performing an initial semantic mining operation on a third adjusted digital education service session log corresponding to the historical digital education service session log to obtain a third initial semantic vector corresponding to the historical digital education service session log, including:

performing semantic aggregation operation on the text semantic relation networks to obtain a third text semantic refining vector corresponding to the historical digital education service session log;

the performing semantic sampling operations of different levels on the third text semantic refining vector to obtain digital education service session text semantics of a first text attribute value and digital education service session text semantics of a second text attribute value corresponding to the historical digital education service session log, including:

9. A data management processing system comprising a processor and a memory; the processor is communicatively connected to the memory, the processor being configured to read a computer program from the memory and execute the computer program to implement the method of any of claims 1-8.

10. A computer readable storage medium, characterized in that a program is stored thereon, which program, when being executed by a processor, implements the method of any of claims 1-8.