WO2022111244A1

WO2022111244A1 - Data processing method and apparatus, electronic device and storage medium

Info

Publication number: WO2022111244A1
Application number: PCT/CN2021/128403
Authority: WO
Inventors: 张祥; 王芷璇; 李丹
Original assignee: 腾讯科技（深圳）有限公司
Priority date: 2020-11-26
Filing date: 2021-11-03
Publication date: 2022-06-02
Also published as: CN112631997B; CN112631997A

Abstract

A data processing method and apparatus, an electronic device and a storage medium. Said method comprises: displaying a test paper creating page; triggering by means of the test paper creating page to acquire test paper data; displaying an online test paper document according to the test paper data; and publishing the online test paper document.

Description

Data processing method, device, electronic device and storage medium

CROSS-REFERENCE TO RELATED APPLICATIONS

This application is based on the Chinese patent application with the application number of 202011347748.0 and the filing date of November 26, 2020, and claims the priority of the Chinese patent application. The entire content of the Chinese patent application is incorporated herein by reference.

technical field

The present application relates to the field of computer technology, and in particular, to a data processing method, a data processing apparatus, an electronic device, and a computer storage medium.

Background technique

With the rapid development of Internet technology, online examination has become a mainstream examination method. Online examination mainly relies on the online examination system. etc.) exam software. The test process of the online test involves: the question maker creates and corrects the test paper through the test software. Correspondingly, the answerer can receive the test paper and complete the test through the test software. Through the investigation, it is found that in the process of creating test papers through these test software, the question-makers often need to set the question types (such as multiple-choice questions, fill-in-the-blank questions, etc.) and format according to the test, which is less flexible.

For this, no effective solution has been provided so far.

SUMMARY OF THE INVENTION

The embodiment of the present application provides a data processing method, which is executed by an electronic device, and the method includes:

Display the test paper creation page;

Trigger the acquisition of test paper data through the test paper creation page, and the test paper data is the data uploaded or input triggered by the test paper creation page;

Display online test paper documents according to test paper data;

Publish online exam papers.

An embodiment of the present application provides a data processing apparatus, and the data processing apparatus includes:

A display unit, configured to display the test paper creation page;

The processing unit is configured to trigger the acquisition of test paper data through the test paper creation page, and the test paper data is the data that is triggered to upload or input through the test paper creation page;

The processing unit is also configured to display the online test paper document according to the test paper data;

Publishing unit, configured to publish online test paper documents.

An embodiment of the present application further provides an electronic device, the electronic device includes a processor and a storage device; the storage device is used to store program instructions; the processor is used to call the program instructions and execute the data processing method provided by the embodiments of the present application .

Embodiments of the present application further provide a computer storage medium, where program instructions are stored in the computer storage medium, and when the program instructions are executed, are used to implement the data processing method provided by the embodiments of the present application.

Embodiments of the present application provide a computer program product or computer program, where the computer program product or computer program includes program instructions, and the program instructions are stored in a computer-readable storage medium. The processor of the electronic device reads the program instruction from the computer-readable storage medium, and the processor executes the program instruction, so that the electronic device executes the data processing method provided by the embodiment of the present application.

Description of drawings

In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the accompanying drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only These are some embodiments of the present application. For those of ordinary skill in the art, other drawings can also be obtained based on these drawings without any creative effort.

1 is a schematic diagram of a data processing system provided by an embodiment of the present application;

2a-2b are schematic flowcharts of a data processing method provided by an embodiment of the present application;

3a to 3g are schematic diagrams of a page provided by an embodiment of the present application;

4a-4b are schematic flowcharts of parsing test paper documents provided by an embodiment of the present application;

5 is a schematic flowchart of another data processing method provided by an embodiment of the present application;

6a to 6c are schematic diagrams of a scenario provided by an embodiment of the present application;

7 is a schematic structural diagram of a dependency tree provided by an embodiment of the present application;

8 is a schematic structural diagram of a feature relationship sub-table provided by an embodiment of the present application;

9 is a schematic structural diagram of a data processing apparatus provided by an embodiment of the present application;

FIG. 10 is a schematic structural diagram of a terminal provided by an embodiment of the present application.

Detailed ways

The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, but not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the protection scope of the present application.

In the following description, reference is made to "some embodiments" which describe a subset of all possible embodiments, but it is understood that "some embodiments" can be the same or a different subset of all possible embodiments, and Can be combined with each other without conflict.

In the following description, the term "first\second\third" is only used to distinguish similar objects, and does not represent a specific ordering of objects. It is understood that "first\second\third" is used in Where permitted, the specific order or sequence may be interchanged to enable the embodiments of the application described herein to be practiced in sequences other than those illustrated or described herein. In the following description, reference to the term "plurality" refers to at least two.

At present, in the process of creating the test paper by the test software, the test object often needs to set according to the question type (such as multiple-choice questions, fill-in-the-blank questions, etc.) and format set by the test software, which is less flexible. The creation of different types of test questions has its own corresponding fixed format, which means that there are different operation processes for the creation of different types of test questions. For users who are not familiar with the operation process, it will greatly reduce the efficiency of test paper creation.

Based on this, an embodiment of the present application proposes a data processing system, and proposes a data processing solution based on the data processing system. Referring to FIG. 1 , the data processing system proposed in the embodiment of the present application may include: a question issuing terminal, a question answering terminal, and an application server. The test object creates the test paper, publishes the test paper, corrects the test paper, etc.; the answering terminal refers to the terminal used by the answering object (that is, the answerer, such as a student), and is used for the answering object to receive the test paper, answer the test paper, submit the answer, and so on. The terminals involved in the embodiments of the present application (such as a question issuing terminal, a question answering terminal, etc.) may be any of the following: a portable device such as a smart phone, a tablet computer, a laptop computer, a desktop computer, and the like.

The target application can be installed and run in the question issuing terminal and the answering terminal. The target application may refer to an application or tool that provides test-related services (such as creating test papers, publishing test papers, correcting test papers, answering test papers, etc.). The tool can be integrated into the application program to provide users with examination-related services, such as a multi-person online collaboration document tool; accordingly, the application server provides corresponding services to the target application (such as automatic judgment based on the answering results returned by the answering terminal). It can be understood as the background server of the target application, and the application server can be an independent server or a server cluster composed of multiple servers. It should be noted that FIG. 1 only exemplarily illustrates the architecture of the data processing system involved in the embodiments of the present application, and does not limit the specific architecture of the data processing system. For example, the data processing system may not include an application server, and the question issuing terminal may directly perform data interaction with the answering terminal.

Based on the above-mentioned data processing system, this application proposes a data processing scheme, which can be executed by the above-mentioned question issuing terminal and answering terminal. As an example, a schematic diagram as shown in FIG. 2 a is provided, which will be described in conjunction with FIG. 2 a :

S10: The question-generating terminal triggers acquisition of test paper data through the test paper creation page, and displays the online test paper document according to the test paper data, wherein the test paper data may be data uploaded or input triggered by the test paper creation page. In some embodiments, when the test paper data is the data that is triggered to be input through the test paper creation page, the test paper data can be understood as data input by the test object in a newly created online document, where the online document in which the test paper data is input is the For the above-mentioned online test paper documents, the above-mentioned online documents and online test paper documents can be understood as: documents that can be edited and viewed online. It can be seen that in this case, the test object can directly complete the creation of the test paper by editing the online document. The created online test paper document is completely generated based on the test paper data input by the test object, and there is no need to limit the test question type and test content during the creation process. The format of filling in is conducive to enhancing the flexibility of test paper creation.

In some embodiments, when the test paper data (for example, a pre-edited word version test paper document) is data triggered to be uploaded through the test paper creation page, the above-mentioned online test paper document parses the test paper data and is generated based on the analysis result. . It can be seen that in this case, the created online test paper document is completely generated based on the test question data uploaded by the test object, and there is no need to limit the test question type and test content filling format during the creation process, which is conducive to enhancing the flexibility of test paper creation; In addition, from the perspective of operation, for the object of the question, the creation of the test paper only needs to upload the test paper data through the test paper creation page, and does not require complex operation procedures to create, which is conducive to greatly improving the efficiency of test paper creation.

S11: The test terminal publishes the online test paper document.

S12: The answering terminal displays the target test paper document associated with the above-mentioned online test paper document, and submits the obtained answer to the above-mentioned target test paper document. Wherein, the content displayed in the online test paper document may include at least one of the following: the test question content, test question score and reference answer of each test question; the target test paper document associated with the above test paper document can be understood as: The same, but does not include the test paper documentation for the reference answers to each test question. The answering object can fill in the answer in the target test paper document displayed on the answering terminal, and the answering terminal can submit (send) the answer to the questioning terminal. The application server automatically scores the answers.

S13: The question-generating terminal displays an answering progress page associated with the online test paper document, and the answering progress page includes answering information and object information of answering subjects who have completed answering the questions. Wherein, the answer information of any answering subject may include: the time when the any answering subject submitted the answer; the predicted total score obtained by the any answering subject answering the above-mentioned target test paper document, and the predicted total score here is for the any one of the answering subjects. The scoring results obtained by the automatic scoring of the answers submitted by the respondents to the above-mentioned target test paper documents can provide a reference for the subsequent corrections of the test papers by the respondents. The above-mentioned automatic scoring can be completed by the application server or the test terminal itself. When the automatic scoring is completed by the application server, the application server can return the scoring result to the test terminal, and the test terminal will display the score. result.

S14, when the question-generating terminal detects a viewing operation input for the target answering object on the answering progress page, display the detailed information of the test paper of the target answering object, and the detailed information of the test paper includes at least one of the following: the reference answer of each test question, the target answering question The content of the answer entered by the subject when answering each test question, and the result of the scoring. The scoring result includes: the predicted total score obtained by the target answering subject for answering the above-mentioned target test paper document, and the predicted answering score obtained by the target answering subject in each test question. The target answering object is any answering object that has completed the answering question.

For example, after the question-generating terminal publishes the online test paper document, the question-generating object can view the answering status of each answering object by triggering the answering progress page when he wants to check the answering status of each answering object (that is, the detailed information of the test paper) or make corrections on the test paper. Take the question object to correct the test paper of the target answer object as an example to illustrate: when the question object wants to correct the test paper of the target answer object, the question object can click the target answer object (the target answer object's logo) on the answering progress page. , so as to trigger the question-generating terminal to display the test paper details information of the target answering object, and the question-generating object can correct the test paper according to the test paper details information.

It can be seen from the above that implementing the embodiment of the present application can preview and publish online test paper documents according to the test paper data, and there is no need to limit the test question type and the filling format of the test question content in the process of creating the test paper, which is beneficial to greatly improve the flexibility of test paper creation.

Based on the description of the above data processing scheme, an embodiment of the present application proposes a data processing method. The data processing method can be executed by the above-mentioned question generating terminal. Please refer to FIG. 2b. The data processing method can include the following steps S201- S203:

S201: Display a test paper creation page, and trigger the acquisition of test paper data through the test paper creation page, where the test paper data is data uploaded or input triggered by the test paper creation page.

Wherein, the above-mentioned test paper creation page may include: a first entry for uploading test paper data (for example, the "import test paper" button in Fig. 3a), and a second entry for creating a new online document (for example, "New test paper" in Fig. 3a ) " button). In some embodiments, when the test paper data is data triggered to be uploaded through the test paper creation page, the test paper data may be an existing electronic document (for example, a word version test paper document), or a paper identified by scanning or photographing quality documents or books, etc. Illustratively, as shown in FIG. 3a, it is assumed that the test paper data is the test paper document 1 pre-stored in the local by the test object. When the test object wants to create a test paper, the "Import test paper" button can be triggered to import the locally stored test paper. Document 1.

In some embodiments, when the test paper data is input data triggered by the test paper creation page, the test paper data can be understood as the data input by the test object in a newly created online document, for example, it may include test paper questions, test content, etc. Wait. Illustratively, as shown in Fig. 3b, when the question-generating object wants to create a test paper, the "New Test Paper" button is triggered, and the question-generating terminal can display the online document as shown in Fig. 3b, and the question-generating object can enter the online document. Exam data. The above method displays the first entry and the second entry in the test paper creation page, so that the subject can quickly create the test paper according to their own habits and needs, improve the flexibility and efficiency of creating the test paper, and avoid the problem that the subject is frequently operating. The resulting waste of computing resources.

S202: Display the online test paper document according to the test paper data.

Here, the question-generating terminal displays the online test paper document according to the test paper data, which is convenient for the test-generating object to preview the online test paper document.

In some embodiments, if the test paper data is data triggered to be uploaded through the test paper creation page (ie, the uploaded test paper document), the online test paper document parses the test paper data and is generated based on the analysis result, and the analysis result includes at least the following: One: the question content, question type, question stem, answer area, reference answer, and target score information of the test paper data of each question in the test paper data.

Assuming that the above test paper data is the uploaded test paper document, the process of parsing the test paper document can be referred to as shown in Figure 4a, including: (1) parsing the content of the document; (2) judging the type of the test question; (3) parsing the stem and answer of the test question (4) Complete the analysis of the test paper documents. in:

Judging the type of test questions: After parsing the document content of the uploaded test paper document, it is necessary to understand the content of the document. First, it is necessary to judge the test question type of each test question in the uploaded test paper document. The test question type can include: single-choice questions, multiple-choice questions , fill-in-the-blank questions, true-false questions, quiz questions, and composition questions (which can also be regarded as a type of quiz questions), etc. An example implementation of judging the test question type is: reading the document content line by line, identifying the first-level serial number in the document and the text content corresponding to the first-level serial number, and matching the text content corresponding to the first-level serial number with the keywords corresponding to each test question type. , and determine the question type of each question according to the matching result.

Among them, the keywords corresponding to each test question type can be as shown in Table 1. The source of these keywords can be pre-entered manually, on the other hand, can be obtained through big data analysis of multiple test paper documents. With the big data analysis of more test paper documents, continuous optimization and The test question types and the keywords corresponding to each test question are updated, so that the implementation of this application can support the processing of more test question types and improve the accuracy of judging the test question types.

Table 1

试题类型Question Type	对应关键字corresponding keyword
单项选择题multiple choice questions	选择题、单项选择题、选择multiple choice, multiple choice, choice
多项选择题multiple choices	多项选择、多、选择multiple choice, multiple, choice
判断题True or False	判断judge
问答题Questions and Answers	问答、作答question and answer

As an example, a schematic diagram of the test paper document as shown in FIG. 3c is provided. In FIG. 3c, the first-level serial numbers include "one, two, three...", and the second-level serial numbers include "1, 2...". Assume that the content corresponding to any one of the first-level serial numbers in the uploaded test paper document (such as the test paper document shown in Figure 3c) (for the convenience of distinction, it is named the target serial number later) is "XXXX", and the keywords corresponding to each test question type are shown in the table As shown in 1, the keywords "multiple choice", "single choice" and "choice" corresponding to the single-choice questions are used to match the content corresponding to the target serial number. If the matching result indicates that the content corresponding to the target serial number matches the keyword "single choice" If the match is successful, it can be determined that the test questions included between the target serial number and the next-level serial number of the target serial number (for example, the test questions corresponding to area 10 in Figure 3c) are of the type: single-choice questions; or, use multiple choice The keywords "multiple choice", "choice" and "multiple" corresponding to the question are matched with the content corresponding to the target serial number. "Select" and "Multi" match successfully, then it can be determined that the type of test questions included between the target serial number and the next-level serial number of the target serial number is: multiple-choice questions; ” is matched with the content corresponding to the target serial number. If the matching result indicates that the content corresponding to the target serial number is successfully matched with the keyword “judgment”, the type of the test questions included between the target serial number and the next-level serial number of the target serial number can be determined. For: judgment questions.

Or, use the quiz question to match the content corresponding to the keyword "question and answer" and "answer" with the target serial number. If the matching result indicates that the content corresponding to the target serial number is successfully matched with the keyword "question and answer" or "answer", it can be determined. The test questions included between the target serial number and the next-level serial number of the target serial number (for example, the test questions corresponding to the area 11 in FIG. 3c ) are of the test question type: quiz questions. After the matching result indicates that the content corresponding to the target serial number is successfully matched with the keyword "question and answer" or "answer", it can also be determined whether there are multiple lines left after the question of the test question included between the target serial number and the next-level serial number of the target serial number White, if there are multiple lines of blank space, the test questions included between the target serial number and the next-level serial number of the target serial number are determined as quiz questions; on the contrary, if there are no multiple lines of blank space, the difference between the target serial number and the target serial number is determined. Questions included between the next level serial numbers are not quiz questions.

Analyze the question stem and answer area of the test questions: According to the above-mentioned test question types, you can continue to analyze the serial numbers of different test questions under the same test question type, and determine the separation between different test questions according to the serial numbers of different test questions. For example, for multiple-choice questions, the content corresponding to each multiple-choice question serial number can be used as the stem of each multiple-choice question. (B)", "a)", "b)" or "(a)", "(b)", the content from each option label to the next option label is each option, and each option is available Options (ie answering area); fill-in-the-blank questions can use the content between the serial numbers of each question as the stem, and identify the blank area with underscores or "()" as the answering area; true-and-true questions can use the content corresponding to the serial numbers of each judgmental question As the stem of each true-and-true question, each sequence of the next level of the stem is the answering judgment option of each true-ordinary question, and each answering judgment option is the answering area of the true-false question.

Completing the analysis of the test paper document: After completing the analysis of the test paper document, on the one hand, a test paper identification can be generated for the uploaded test paper document, and the analysis result obtained by the above analysis can be stored in the target storage area in association with the test paper identification. The analysis result of the test paper document is obtained, and the analysis result includes the test question content, test question type, question stem and answering area of each test question in the test paper document. Wherein, the above-mentioned target storage area may include a local storage area, a blockchain, a cloud server or an application server, which is not specifically limited.

On the other hand, an online test paper document can be generated based on the analysis results, and the implementation process can be as follows: according to the test question type, question stem and answering area of each test question in the test paper document, the content of each test question is integrated and processed to obtain an online test paper document, and output The question terminal can display the online test paper document (the display effect can be as shown in Figure 3a), and complete the preview of the online test paper document. Through the above method, the analysis of the test paper document can be realized intelligently and automatically, and the online test paper document can be accurately generated without user operation during the period, which can effectively save the labor cost, and can also save the computing resources consumed by the electronic device for responding to the user operation.

It can be understood that, in some embodiments, the analysis of the uploaded test paper document and the generation of the online test paper document may be performed simultaneously, that is, the online test paper document is generated while parsing. For example, when the question-generating terminal reads the document content of the test paper document line by line, it can first determine the first level included between the first level number and the second level number according to the text content corresponding to the first level number. Types of questions in the first type of questions, identify the question stems and answering areas of each question in the first type of questions, so as to integrate the first type of questions according to the type of questions in the first type of questions, and the stems and answering areas of each question in the first type of questions The content of the test questions in each test question, including the test question content of the first type of test questions and the online test paper document of the answering area, and so on, you can continue to analyze the second level between the second level serial number and the third level number. Types of questions, stems, answering areas, etc. of the similar questions, and update the content of the online test paper document, until the analysis of all the content in the test paper document is completed, then stop. Alternatively, in some embodiments, after all contents of the test paper document are parsed, an online test paper document may be generated based on the analysis result, which is not specifically limited in this embodiment of the present application.

From the above content, it can be seen that the above uploaded test paper documents do not contain the reference answers and score-related information of the test questions (for example, "total xx questions", "total xx points", "each question xx" and other information marked after each type of question type. ), in some embodiments, the uploaded test paper document may include reference answers and score-related information of the test question, and the process of parsing the test paper document may refer to Figure 4b, including: (1) parsing the content of the document; (2) Judging the type of test question; (3) Analyzing the question stem and answering area of the test question; (4) Analyzing the reference answer of the test question; (5) Analyzing the score of the test question, (6) Completing the analysis of the test paper document. Among them, the implementation of parsing the document content, judging the type of test questions, analyzing the question stem and answering area of the test question can refer to the above content, and the implementation of other parts can refer to the following description:

Parsing the test question answer: For example, after parsing the test question type and the question stem, the reference answer of the test question can be analyzed at the target position corresponding to each test question, wherein the target position is associated with the test question type of the test question. For example, before and after the stem of a multiple-choice question, there can be "(A)", "(AB)" or "[A]", "[AB]" to mark the reference options of multiple-choice questions. Correspondingly, multiple-choice questions The corresponding target position is the position corresponding to the reference options before and after the stem of the multiple-choice question, and the reference answer of the multiple-choice question is "(A)", "(AB)" or "[A]", "[AB] before and after the stem of the question. " and other marking information; after the option of the true-and-true question, you can add "(yes)", "(no)", "(correct)", "(wrong)" and other judgment marks. Accordingly, the target position corresponding to the true-false question is In order to determine the corresponding position of the judgment mark after the judgment question option, the reference answer of the judgment question is the judgment mark such as "(yes)", "(no)", "(correct)", "(wrong)" after the judgment question option; fill in the blank Questions generally have formats such as underline, "()", "[]", etc. Correspondingly, the target position corresponding to the fill-in-the-blank question is the position corresponding to the format such as underline, "()", and "[]". The reference answer of the fill-in-the-blank question is Underscore, "()", "[]" and other formats; the parsing rules of the reference answer of the question and answer question can be the same as that of the fill-in-the-blank question.

Analysis of test question scores: For example, each type of question is usually marked with "total xx questions", "total xx points", "each question xx" and other score-related information. Obtain the score-related information of each type of question type from the document content, and analyze the score-related information corresponding to each type of question type to obtain the question score of each question in the test paper document, so as to complete the analysis of the test question score, and get Target score information, the target score information includes: the score-related information marked after each type of question, and the question score of each test question.

Complete the analysis of the test paper: After completing the analysis of the test paper, you can get the above target score information, the test content of each test question, the test question type, the question stem, the answer area, and the reference answer. That is, when the uploaded test paper document contains the reference answers and score-related information of the test question, the test paper document is parsed, and the obtained analysis result may include: target score information, test question content of each test question, test question type, question stem, Answer area and reference answers. In some embodiments, the question-generating terminal may perform similar steps to those performed after completing the test paper analysis in FIG. 4a: on the one hand, a test paper identification may be generated for the uploaded test paper document, and the above-mentioned analysis result and the test paper identification may be associated and stored in the target storage The above analysis results can be obtained directly according to the test paper identification. On the other hand, an online test paper document can be generated and displayed based on the analysis results (the display effect can be shown in Figure 3d). The online test paper document is marked with a reference answer for each test question. Question", "total xx points", "each question xx" and other score-related information. In the above manner, in the case where the uploaded test paper document contains the reference answers and score-related information of the test question, the corresponding analysis can be performed accurately and effectively to ensure the comprehensiveness and accuracy of the obtained online test paper document.

In some embodiments, when the test paper data is the data that is triggered to be input through the test paper creation page, the test paper data can be understood as data input by the test object in a newly created online document, where the online document in which the test paper data is input is the It is the above online test paper document, so as to realize the preview of the online test paper document. The test paper data may include the content of the test questions, reference answers, target score information, and the like. In some embodiments, after the creation of the online test paper document is completed, the question-generating terminal may generate a test paper ID (for example, a test paper ID), and store the inputted test paper data in association with the test question ID in the target storage area. Subsequently, the test paper ID can be directly stored according to the test paper ID. Obtain the above test paper data.

Among them, in some embodiments, for the data stored in the target storage area (such as the above-mentioned analysis results or test paper data), a read permission can also be set. The read permission of the answer and target score information. The answer object of the test paper has the read permission of the test question content and the target score information in the target storage area, and other objects do not have the read permission to any information in the target storage area. In this way, the data security of the target storage area can be improved.

203: Publish online test paper documentation.

For example, the online test paper document can be displayed on the test paper editing page (for example, as shown in Figure 3a). The test paper editing page includes a release entry, and the question object can input a trigger operation for the release entry by clicking, pressing or voice, etc., and the question issuing terminal After detecting the trigger action, the online test paper document can be published.

For the test object, after viewing the online test paper document, you can modify any part of the test paper, including the content of the test question, test question scores and reference answers, etc. Correspondingly, you can modify any part of the test paper based on the test object. The above-mentioned analysis result or test paper data pre-stored in the target storage area, for example, modify the test question content, test question score, reference answer, etc. of any test question pre-stored in the target storage area. For example, before publishing the online test paper document, the question-generating terminal may perform editing processing on the online test paper document according to the editing operation input for the online test paper document, and the editing process is used to edit at least one of the following: the content of the target test question in the online test paper document , test scores and reference answers. In some embodiments, pre-stored related data in the target storage area may be modified based on the editing result.

In some embodiments, the editing process may be used to edit the content of the target question in the online test paper document, which may include any content of the question, such as stem content, options, sequence numbers, and the like. Illustratively, referring to Figure 3e, when the question maker views the online test paper document, and wants to change the B option of the first multiple-choice question (that is, the target test question) from "Water Huang" to "Locusts", he can directly in the Click the display position corresponding to option B in the online test paper document, and change "Water Huang" to "locust", so as to complete the editing of the content of the first multiple-choice question. In this case, the operation of clicking option B and changing "Water Huang" to "locust" is the above editing process.

In some embodiments, the editing process may be used to edit question scores and/or reference answers for target questions in an online test paper document. The online test paper document includes the test content and answering area of each test question. An example implementation of editing and processing the online test paper document by the question-generating terminal according to the input editing operation for the online test paper document may be: when detecting a triggering operation for the answering area corresponding to the target test question (for example, clicking or pressing the corresponding answering area for the target test question) When the answer editing entry of the target question is displayed, the answer editing area is triggered to be displayed through the answer editing entry, and the target information entered for the target question is received in the answer editing area, and the target information includes at least one of the following: the reference answer of the target question and the Exam score. In some embodiments, at least a portion of the objective information may be displayed in the online test paper document.

Illustratively, referring to Figures 3f and 3g, assuming that the target test question is the question and answer question in Figure 3f "Briefly describe photosynthesis and write the corresponding reaction formula", after the question-producing subject checks the online test paper document, if they want to set the question and answer You can click the answering area of the question to trigger the answering terminal to display the answer editing entry (the "Set Answer" button included in the upper picture in Figure 3f), and the question object clicks the answer editing entry to trigger the question. The terminal displays the answer editing area (as shown in the lower figure in Figure 3f), and the question object can enter the reference answer and test question score corresponding to the target question in the answer editing area, so as to complete the editing processing of the test question score and reference answer. In some embodiments, after the editing of the reference answer and the test question score corresponding to the target test question is completed, the question-generating terminal can update and display the reference answer and test question score corresponding to the target test question in the online test paper document, and the display effect is shown in Figure 3g. . Or, after editing the reference answer and test question score corresponding to the target test question, the test terminal can update and display only the test question score of the target test question in the online test paper document, and display the reference answer and test question corresponding to the target test question entered by the test object. Scores are stored in the target storage area for subsequent retrieval. Among them, the reference answer entered in the answer editing area can be text (Chinese, English or other text), or you can enter a formula. The input method of the formula can include: copying and pasting the formula, calling the formula editor to edit, or inputting through the input method, This is not specifically limited. Through the above-mentioned editing process, the accuracy of the online test paper document can be improved, and the waste of computing resources caused by re-creating the test paper can be avoided.

In the embodiment of the present application, a test paper creation page may be displayed, the test paper data is triggered to be acquired through the test paper creation page, and the online test paper document is displayed according to the test paper data, thereby publishing the online test paper document. In the process of creating the test paper, there is no need to limit the type of test questions and the filling format of the test content, which is beneficial to greatly improve the flexibility of test paper creation.

Referring to FIG. 5 , the implementation of the present application proposes a data processing method. The data processing method can be executed by the above-mentioned question generating terminal, and the data processing method can include the following steps S501-S504:

S501: Display the test paper creation page, and trigger the acquisition of test paper data through the test paper creation page.

S502: Display the online test paper document on the test paper editing page according to the test paper data. For the implementation of steps S501 to S502, reference may be made to the relevant descriptions of steps S201 to S202 in the foregoing embodiments.

S503: In response to the triggering operation for the release portal, generate a test paper link associated with the online test paper document. Wherein, the test paper link includes the test paper identification of the online test paper document, for example, the test paper ID.

S504: Share the test paper link to the answering terminal, where the test paper link is used for: the answering terminal triggers the acquisition of the target test paper document matching the online test paper document through the test paper link, so that the respondent can answer the target test paper document through the answering terminal.

In some embodiments, before generating the test paper link associated with the online test paper document in response to the triggering operation for the publishing portal, the question generating object may further set the total answering time, the answering start time, and enable the function of timed roll-up through the question generating terminal. , setting answer permissions, etc., to obtain answer configuration information for the online test paper document, and store the answer configuration information in the target storage area in association with the test paper ID corresponding to the online test paper document. The answering configuration information includes any one or more of the following: total answering time, answering start time, timed roll-up function indication information, and answering authority information, where the timed roll-up function indication information is used to indicate whether it has been opened for the online test paper document The function of timed rewinding, the answering authority information is used to indicate what kind of object has the answering authority for the online test paper document. For the question-generating object, any object can be set to have the right to answer questions. In this case, the above-mentioned question-answering permission information can indicate that any object has the right to answer questions for the online test paper document; Answering authority, for example, the question issuing object can set the answering object list through the question issuing terminal, and the answering object in the answering object list has the answering authority. In this case, the above answering authority information includes the answering object list, and is used to indicate the answering object list The answering object of , has the answering permission for the above-mentioned online test paper document.

In some embodiments, the test paper editing page further includes an answer configuration entry (for example, the “Settings” button shown in FIG. 3a ), and the test terminal can trigger the answer configuration entry to trigger the answer configuration window, and set in the answer configuration window The total answering time, answering start time, enabling the function of timed rewinding, setting answering permissions, etc., can obtain the answering configuration information for the online test paper document to complete the answering configuration of the online test paper document. In some embodiments, after completing the answer configuration of the online test paper document, the answer configuration information may be stored in a target storage area in association with the test paper identification of the online test paper document, so as to facilitate subsequent acquisition. Through the above-mentioned answering configuration method, the answering process can be effectively constrained, and the waste of manpower and computing resources caused by invalid answering can be avoided.

In some embodiments, after generating the test paper link associated with the online test paper document in response to the triggering operation for the publishing portal, the test paper terminal can display the test paper link and the share button, and trigger the determination of the target sharing method and target sharing address through the share button, And according to the target sharing method and target sharing address, the test paper link is shared to the answering terminal, so as to complete the distribution of the test paper. The target sharing method may include: third-party application sharing, group sharing, photo album sharing, email sharing, etc. The third-party application may refer to a social application, such as an instant messaging social application.

Exemplarily, referring to Fig. 6a, when the question-generating object wants to publish the online test paper document, a triggering operation can be input for the publishing entry by clicking, pressing or voice, and after detecting the triggering operation, the question-generating terminal can generate and display the The test paper link and share button associated with the online test paper document, and the share button is used to trigger the test terminal to display the sharing method selection list. When the subject of the question clicks the button corresponding to "Share to third-party application" in the sharing method selection list, the third-party application sharing can be determined as the target sharing method. In some embodiments, the target sharing address can be determined. The target sharing address It can be any one or more groups, friends, or content sharing platforms in the third-party application for the subject of the question.

When the subject of the question clicks the button corresponding to "Share to my group" in the sharing method selection list, the group sharing can be determined as the target sharing method. In some embodiments, the subject of the question can select his Target group, the question-generating terminal can determine the target group as the target sharing address.

When the test object clicks the button corresponding to "Save to Album" in the sharing method selection list, the album sharing can be determined as the target sharing method, and the test terminal can generate an image including the link of the test paper, and add the image to the local photo album. Subsequently, the subject of the question can share the image to any target sharing address that he or she wants to share.

When the subject who creates the question clicks the button corresponding to "Send by mail" in the sharing method selection list, the e-mail sharing can be determined as the target sharing method. The one or more email addresses may be determined as target sharing addresses.

Alternatively, in some embodiments, in the case that the question-generating object has preset the above-mentioned list of answering objects, the question-generating terminal, in response to the triggering operation for the publishing portal, generates the test paper link associated with the online test paper document, and then directly sends the test paper to the test paper. The link is sent to the answering terminal corresponding to the answering object in the answering object list, so as to complete the automatic distribution of the test paper. Through the above method of sharing the test paper link, the answering terminal can accurately obtain the target test paper document through the test paper link and answer the question. At the same time, because the test paper link itself is relatively light, the sharing efficiency and success rate can be improved; by determining the target sharing method Sharing the address with the target can improve the flexibility and accuracy of sharing, and at the same time ensure the security of the sharing link and the online test paper document.

After the test paper distribution is completed, the recipient receives and clicks the above-mentioned test paper link (the test paper link includes the test paper identifier) through the answering terminal. In some embodiments, the answering terminal can verify whether the receiving object has the above-mentioned online test paper document. If it does not exist, the test paper generation information (including: the above-mentioned target score information corresponding to the online test paper document, the test question content of each test question in the online test paper document, etc. etc.), and based on the test paper generation information, a new test paper document (ie, a test paper document with blank answers) is generated, and the new test paper document is the target test paper document matching the online test paper document. In some embodiments, an area authority control can be generated for the target test paper document, which is used to restrict the answering area for modifying or filling in the answer of the test question, and also to limit the answering authority. Afterwards, when the recipient answers the target test paper document, they can only modify or fill in the test questions in the answering area, and only the recipients who have the permission to answer the question can modify or fill in the answer to the test question in the answering area.

Or, if the answering terminal detects that the receiving object has the answering record of the above-mentioned online test paper document, it can detect whether the receiving object has completed the answering of the online test paper document (that is, whether it has been handed in), and if it has been completed, the output is used to prompt the answer. Completed prompt message. Alternatively, if the recipient has not completed the answer to the online test paper document, a record of the recipient's answer to the online test paper document can be obtained, and a target test paper document is generated based on the answer record, and the subsequent recipient can continue to answer the target test paper document.

Wherein, whether the receiving object has the answering authority can be determined according to the above-mentioned answering authority information obtained by the answering configuration of the question-producing object in advance, and the answering authority information is used to indicate what kind of object has the answering authority for the online test paper document, for example, the answering authority information It is used to indicate that the answering object in the answering object list has the answering authority for the above-mentioned online test paper document. It is understandable that, since the online test paper document and the target test paper document essentially belong to the same test paper, the answering authority for the online test paper document can be equivalent to the answering authority for the target test paper document. Assuming that the answering authority information is used to indicate that the answering object in the answering object list has the answering authority for the above-mentioned online test paper document, then, in the subsequent process of the receiving object answering the target test paper document, when the answering terminal detects that the receiving object is directed against the target test paper. During the input operation of the answering area in the document, it can be verified whether the receiving object is any answering object in the answering object list. If so, it is determined that the receiving object has the answering authority. In some embodiments, in response to the input operation, input or modify Answers to the test questions. Conversely, if the receiving object is not any answering object in the answering object list, it is determined that the receiving object does not have the right to answer the question, and accordingly, the answering terminal may not respond to the above input operation. It is prompted that the recipient does not have permission to answer the question. In this embodiment of the present application, the receiving objects that have the right to answer questions to the target test paper document may be collectively referred to as answering objects.

Among them, the generation of the target test paper document and the regional authority control can be generated by the application server (for example, the application server in FIG. 1 ) in addition to the answering terminal itself, and the generated target test paper document with the regional authority control is issued to The answering terminal is displayed by the answering terminal, and the answering object can answer based on the displayed target test document. The server or the question issuing terminal may record the time when the answer result is received, and record the time when the answer result is received as the time when the answer object submits the answer (hereinafter referred to as the answer submission time). In addition, the application server or the question-generating terminal can automatically judge the answer result, and obtain the scoring result, and the scoring result includes: the predicted total score obtained by the answering object for answering the above-mentioned target test paper document, and the target answering object in the The predicted answer score for each question.

It can be understood that when the above steps of recording the answer submission time and automatic scoring are performed by the application server, the application server can return the answer submission time and scoring result to the question-generating terminal, so that the subsequent question-generating terminal can display the answer. Submission time and scoring results.

In some embodiments, it is assumed that the question-generating object has set the answering start time before publishing the online test paper document. Then, when the answering terminal detects that the receiving object does not have the answering record of the above-mentioned online test paper document, it can also detect whether the current time is late At the above answering start time, if yes, the subsequent steps related to obtaining the target test paper document will not be performed. On the contrary, a timeout prompt message may be output to remind the recipient that the answering start time has been missed.

Or, assuming that the test object has enabled the timed roll-up function and set the total answering time before publishing the online test paper document, the answering terminal can start the timer when displaying the target test paper document. When the time duration of the timer is equal to When the above total answering time is used, if it is detected that the answering subject has not completed the answering question, the answering result of the answering subject for the target test paper document can be obtained, and the answering subject can be marked as having completed the answering question for the target test paper document, so as to complete the automatic collection. After marking that the answering object has completed the answering questions for the target test paper document, the answering object's answering authority for the target test paper document can be collected, that is, the answering object can no longer modify or fill in the test question answer in the answering area after submitting the paper.

S505: Display an answering progress page associated with the online test paper document, where the answering progress page includes answering information and object information of answering subjects who have completed answering the questions. The object information of any answering object may be the identity information of any answering object, such as a name or account information (such as a nickname) in the target application, and the like. The answer information of any answer object may include: the answer submission time of the any object, and the predicted total score obtained by answering the above-mentioned target test paper document. Exemplarily, for the display effect of the answering progress page, please refer to the answering progress page in FIG. 6b.

S506: In response to the viewing operation for the target answering object in the answering progress page, display the detailed information of the test paper of the target answering object, where the detailed information of the test paper includes at least one of the following: the reference answer of each test question, the input of the target answering object when answering each test question and the scoring result determined according to the answering content, the scoring result includes the predicted answer score obtained by the target answering object in each test question.

For example, after the test subject publishes the online test paper document through the test terminal, and wants to view the test paper details of the target answerer, he can view the test paper details of the target answerer by triggering the answering progress page. If the test object finds that the scoring result for the target question is wrong, it can modify the scoring result. Exemplarily, as shown in Fig. 6b, when the test object wants to view the test paper details of Zhang XX (that is, the target answering object), he or she can click on "Zhang XX" to trigger the test terminal to display the test paper details of Zhang XX. . Clicking "Zhang XX" here is the above-mentioned viewing operation for the input of the target answering object. Through the above method, it is possible to support the test object to view the detailed information of the test paper online, meet the marking requirements of the test object, and improve the data processing efficiency (such as the marking efficiency).

In the embodiment of the present application, the test questions included in the test paper can be divided into two categories: subjective questions and objective questions. The objective questions refer to the test questions with fixed answers, and the scoring of the objective questions can completely avoid the subjective factors of the examiner. Interference, such as true and false questions, multiple choice questions, etc. are all objective questions. In the embodiment of the present application, the method of determining the prediction score of the answer obtained by the target answering object in the objective question may be: comparing whether the answer content input by the answering object is consistent with the reference answer of the objective question, and if they are consistent, mark the answer as correct; marked as wrong answer. All objective questions will be counted, correct questions will be scored according to the score set for the corresponding objective questions, wrong questions will be scored as 0 points or corresponding deductions will be recorded according to the settings.

Subjective questions refer to test questions that do not have a fixed answer. Because the answer to subjective questions is generally organized by the reference test personnel, the answers of the test sites often have different words and terms and different sequences. Some words are added or deleted, but their meanings are different. However, it is the same, and there may be a possibility that when some words are different, the meanings are completely opposite. Due to the flexible and changeable characteristics of the answers to the subjective questions, it is difficult to accurately judge the subjective questions.

At present, the commonly used methods for judging subjective questions are as follows: Method 1: Manual judging by the judges, which is very labor-intensive; If there are key words in the submitted answers, the corresponding points will be awarded, and the results obtained by this method are not accurate enough.

Based on this, the embodiments of the present application can identify the answer content submitted by the answering object for the subjective question by means of formula identification and/or semantic identification, and more accurately determine the scoring result of the subjective question (ie, the predicted answer score), and improve the electronic equipment. The actual utilization rate of the computing resources consumed in the scoring process is also conducive to improving the scoring efficiency of the question-producing objects. For example, the method of determining the prediction score obtained by the target answerer in the subjective question may include: obtaining the target answer content input by the target answerer to answer the subjective question, and the target reference answer of the subjective question, when the target reference answer includes a reference formula , perform formula identification on the formula in the target answer content, obtain the formula identification result, and determine the answer prediction score obtained by the target answer object in the subjective question according to the formula identification result; or, when the target reference answer includes reference text information, the target answer The text information in the answer content is semantically identified, and the semantic identification result is obtained, and the answer prediction score obtained by the target answer object in the subjective question is determined according to the semantic identification result.

It can be understood that, in the embodiment of the present application, if the target reference answer only includes the reference formula, the answering terminal can directly determine the answer prediction score obtained by the target answer object in the subjective question based on the formula recognition result; Including reference text information, the answering terminal can directly determine the answer prediction score obtained by the target answering object in the subjective question based on the semantic recognition result; if the target reference answer includes reference formula and reference text information, then the answering terminal can be based on the formula recognition result and semantics. The recognition result determines the answer prediction score obtained by the target answering object in the subjective question.

Exemplarily, assuming that a certain subjective question is the first question and answer question shown in Figure 3d, "Briefly describe photosynthesis and write the corresponding reaction formula", the total score of this question and answer question is 10 points, and the target reference answer of the question and answer question is given by The reference formula and the reference text information are composed of two parts. The reference score of the reference formula and the reference text information can be preset. Suppose the reference score of the reference formula part is 2 points, and the reference score corresponding to the reference text information part is 8 points. If the score of the target answering object in the formula part is determined to be 2 points based on the formula recognition result, and the score obtained by the target answering object in the text information part is determined based on the semantic recognition result is 5 points, then the target answering object can be placed in the first question. The obtained answer prediction score was determined to be 7 points.

Wherein, the implementation of semantic recognition for the text information in the target answer content includes: performing word segmentation processing on the reference text information in the target reference answer to obtain at least one reference word segment, and when the at least one reference word segment is a keyword of the same sex, Keyword matching is performed on the text information in the target answer content to obtain a semantic recognition result; when any reference participle in at least one reference participle is not a same-sex keyword, and the number of historical reference scoring data existing in the subjective question satisfies the quantitative condition, The text information in the target answer content is judged and scored, and the semantic recognition result is obtained; when any reference participle in at least one reference participle is not a same-sex keyword, and the number of historical reference scoring data existing in the subjective question does not meet the quantitative condition When , the text information in the target answer content is analyzed by dependency syntax, and the semantic recognition result is obtained. In this way, the manner of semantic recognition is determined according to the situation of the reference word segmentation in the reference text information, which can improve the accuracy of semantic recognition. Wherein, same-sex keywords refer to keywords with the same part of speech, for example, they are nouns, verbs, etc. In this embodiment of the present application, the reference text information in the target reference answer may be subjected to word segmentation processing, and after obtaining at least one reference word segmentation, Each reference participle is tagged with part of speech. If it is determined based on the part of speech tagging result that the part of speech of at least one reference participle is the same, at least one reference participle can be determined as a keyword of the same sex. Correspondingly, if it is determined based on the part-of-speech tagging result that there is any part of speech in the at least one reference participle that is different from the part of speech of other reference participles, it can be determined that any one of the reference participles in the at least one reference participle is not a same-sex keyword. The quantity of the existing historical reference scoring data satisfies the quantity condition may be, for example, that the quantity of the existing historical reference scoring data is greater than or equal to the quantity threshold.

Wherein, after performing word segmentation on the reference text information in the target reference answer and obtaining at least one reference word segment, if it is detected that at least one reference word segment is a keyword of the same sex, it is also possible to detect whether the above at least one reference word segment can form a complete sentence , if a complete sentence cannot be formed, the above at least one reference word segment can be processed as a reference keyword sequence (the reference keyword sequence includes at least one reference word segment), and the execution of the above-mentioned keyword matching is triggered to match the text information in the target answer content The steps of performing semantic recognition and obtaining the result of semantic recognition. Here, the implementation of keyword matching on the text information to obtain the semantic recognition result is as follows: performing word segmentation processing on the text information in the target answer content, obtaining at least one word segmentation, and processing the at least one word segmentation into a keyword sequence (the keyword The sequence includes at least one participle), compare (match) the keyword sequence with the above-mentioned reference keyword sequence, and determine the comparison result (matching result) as the semantic recognition result, where the semantic recognition result indicates that the keyword sequence and the reference key The number of successfully segmented word sequences matched. The above sequence-based order matching method is suitable for a scene where the reference word segmentation itself has an order, which can improve the accuracy of semantic recognition in this scene. In some embodiments, the answer prediction score obtained by the target answer object on the subjective question may be determined according to the semantic recognition result.

Exemplarily, assuming that the subjective question is a question-and-answer question, the score of the question set for this question is 4 points, the reference answer only includes reference text information, and the reference keyword sequence corresponding to the reference text information is {cultivation, erudition, truth-seeking, innovation} , the target answer content entered by the target answerer for the question and answer question is "cultivation, erudition, and truth-seeking", and the keyword sequence corresponding to the target answer content is {cultivation, erudition, truth-seeking}, and the keyword sequence {xiude, erudition, truth-seeking} is compared with the above reference keyword sequence {xiude, erudition, truth-seeking, innovation}, and the comparison result indicates that the keyword sequence {xiuzhi, erudition, truth-seeking} is compared with the reference keyword sequence {xiude, erudition, truth-seeking} , Innovation} The number of successfully matched word segmentation is 3, and each successful match is scored as 1 point, then the answer prediction score obtained by the target answering object in the above question and answer question can be determined as 3 points.

In some embodiments, performing keyword matching on the text information in the target answer content to obtain a semantic recognition result includes: performing word segmentation processing on the text information in the target answer content to obtain at least one word segment, and combining the at least one word segment with the above at least one word segment A reference participle performs unordered matching, and the matching result is determined as the semantic recognition result. The semantic recognition result here is used to indicate the number of word segments that successfully match the reference word segment in at least one segment. Exemplarily, suppose that the subjective question is a question and answer question as shown in Figure 6c, and the target reference answer of the question and answer question only includes the reference text information "prosperity, democracy, civilization, harmony, freedom, equality, justice, rule of law, patriotism, dedication, Integrity and friendliness”, at least one reference participle corresponding to the reference text information includes: prosperity, democracy, civilization, harmony, freedom, equality, justice, rule of law, patriotism, dedication, integrity and friendliness. The test score for this question is 12 points. The target answer content entered by the respondents for this question and answer question is "prosperity, democracy, civilization, harmony, freedom, equality, justice, rule of law, patriotism, dedication, and friendliness", that is, the target answer content only includes text information. In this case, the text information can be segmented to obtain at least one segment, the at least one segment includes: prosperity, democracy, civilization, harmony, freedom, equality, justice, rule of law, patriotism, dedication, and friendliness. Further, each participle can be matched with each reference participle in disorder, and the matching result indicates that the number of participles successfully matched with each reference participle in each participle is 11, and each successful match is scored 1 point, then the target answer object can be placed in The answer prediction score obtained for the question and answer question shown in Figure 6c is determined to be 11 points. The above-mentioned out-of-order matching method is suitable for a scenario where the reference word segmentation itself does not have an order (or the order is not important), and can improve the accuracy of semantic recognition in this scenario.

In some embodiments, the text information in the target answer content is subjected to scoring processing to obtain the semantic recognition result. The implementation manner may be: calling the target scoring model to perform scoring processing on the text information in the target answer content to obtain the subjective question The scoring result of the subjective question is determined as the semantic recognition result, and the scoring result of the subjective question is used to indicate: the proportion of the scores of the test questions of the subjective question. Exemplarily, it is assumed that in the process of answering this question, the pre-set test score for any subjective question A is 20 points, the target reference answer of subjective question A only includes reference text information, and the subjective question scoring result output by the target scoring model. Indication: the proportion of the score of the subjective question is 100%, then it can be determined that the predicted score of the answer obtained in this subjective question A is 20 points. The above method is implemented based on the principle of artificial intelligence, which can improve the accuracy of semantic recognition.

In some embodiments, before calling the target scoring model to perform scoring processing on the text information in the target answer content, the initial scoring model may be trained by using historical reference scoring data of subjective questions to obtain the target scoring model, which is The historical reference scoring data includes the historical answer content input by other respondents when answering the above-mentioned subjective question and the historical scoring result of scoring (eg, manual scoring) based on the historical answer content. The initial scoring model may be a latent Dirichlet Allocation (LDA) model, or other neural network models. The above method can improve the accuracy of scoring through the target scoring model by performing model training.

In some embodiments, performing dependency syntax analysis on the text information in the target answer content to obtain a semantic recognition result may include: identifying sentences included in the text information in the target answer content, and performing dependency analysis on the sentences to obtain the text The dependency feature table corresponding to the information. Obtain the reference dependency feature table corresponding to the reference text information, compare the dependency feature table and the reference dependency feature table, and determine the semantic recognition result from the comparison result. The semantic recognition result here is used to indicate the dependency feature table and the reference dependency relationship. The overall match between feature tables. Exemplarily, assuming that the target reference answer of any subjective question only includes reference text information, and the question score set for this question is 12 points, if the semantic recognition result indicates the total matching degree between the dependency feature table and the reference dependency feature table is 50%, then, the predicted score of the answer obtained by the target answering object in any subjective question can be determined to be 6 points. The above method performs semantic recognition based on the dependency relationship of sentences, which can effectively analyze the actual semantics and improve the accuracy of semantic recognition.

A piece of text information may include multiple sentences, each sentence may correspond to a dependency feature sub-table, and the dependency feature table corresponding to the text information may include dependency feature sub-tables corresponding to all sentences in the text information.

In some embodiments, an exemplary way of determining the dependency feature sub-table corresponding to any sentence may be: performing a dependency relationship analysis on any sentence, determining the part of speech, role classification and dependency relationship of each participle in the any sentence, and A dependency relation tree corresponding to any sentence is generated according to the part of speech, role classification and dependency relation of each participle, and then a dependency relation feature sub-table corresponding to any sentence is determined based on the dependency relation tree. Exemplarily, assuming that any sentence in the text information is "Enhancing cooperation between the Export-Import Bank of Country X and Bank of Country X", the dependency tree corresponding to any sentence can be shown in Figure 7, where "enhanced" in Figure 7 is: The predicate is the root (ROOT), other participles are nodes under the root, ARG represents the predicate role, specifically, "ARG=A0" means that the corresponding participle "bank" is the previous word of the predicate "strengthen", "ARG=A1 " indicates that the corresponding participle "cooperation" is the last word of the predicate "strengthening"; W indicates a word; R indicates a dependency relationship, specifically, "R=SBJ" indicates a subject-predicate relationship, and "R=NMOD" indicates a dependency relationship For the fixed relationship, "R=COMP" indicates that the dependency relationship is a verb-object relationship; G represents the part of speech, including: VV (verb), NN (other nouns), NR (proper nouns), CC (coordinated conjunctions), etc. . The dependency feature sub-table shown in FIG. 8 can be obtained according to the dependency tree. The above method effectively expresses the dependencies in the sentence through the dependency tree, thereby ensuring the accuracy of the obtained dependency feature sub-table and the dependency feature table, and facilitating accurate and effective semantic recognition.

The reference dependency feature table corresponding to the reference text information is pre-generated, and the reference dependency feature table may include a reference dependency feature sub-table corresponding to each reference sentence in the reference text information. The generation of each reference dependency feature sub-table The way is similar to the way of generating the dependency feature sub-table above.

In some embodiments, the dependency feature sub-table of each statement in the text information may be sequentially compared with the reference dependency feature sub-table of each reference statement in the reference text information, for example: the dependency feature sub-table of the first statement Compare the table with the reference dependency feature sub-table of the first reference statement, compare the dependency feature sub-table of the second statement with the reference dependency feature sub-table of the second reference statement, and so on until the comparison is complete The last statement in the text message. Based on the comparison results of each reference dependency feature sub-table and each dependency feature sub-table, the matching degree between each reference dependency feature sub-table and each dependency feature sub-table can be determined, and then the matching degrees are summed to obtain The total matching degree between the dependency feature table corresponding to the text information and the reference dependency feature table corresponding to the reference text information is taken as the semantic recognition result. The above method compares sentences in units, and sums up the obtained matching degrees to obtain a total matching degree, so that the total matching degree can effectively reflect the relationship between text information and reference text information, that is, it can effectively improve the accuracy of semantic recognition. .

In some embodiments, the number M of reference sentences in the reference text information may be predetermined (M is an integer greater than 0), and a corresponding reference matching degree may be allocated to each reference sentence, for example, it may be evenly allocated, for example, each reference sentence may be allocated The same reference matching degree (100/M)%; or according to the importance of each reference sentence, each reference sentence is assigned a different reference matching degree, but the sum of all reference matching degrees needs to be 100%.

In some embodiments, the dependency feature sub-table of each statement in the text information may be sequentially compared with the reference dependency feature sub-table of each reference statement in the reference text information. The reference dependency feature sub-table of any reference sentence is successfully matched, then the total matching degree between the dependency feature table corresponding to the text information and the reference dependency feature table corresponding to the reference text information (the initial value of the total matching degree is 0) Accumulate the reference matching degree corresponding to any reference sentence until the last sentence in the text information is compared to obtain the latest total matching degree. That is, if it is detected that the dependency feature table of any statement and the reference dependency feature sub-table of any reference statement fail to match, determine the dependency feature table of any statement and the reference dependency feature of any reference statement. The matching degree between the subtables is zero.

For example, reference text information includes reference sentence 1, reference sentence 2, and reference sentence 3, and the reference matching degrees set for each reference sentence are shown in Table 2. The text information includes: sentence 1, sentence 2, and sentence 3. In this case, according to the sorting order of the sentences in the text information, the dependency feature table of sentence 1 can be compared with the reference dependency features corresponding to reference sentence 1, reference sentence 2 and reference sentence 3, respectively, and the sentence can be obtained by comparison. The dependency feature table of 1 and the reference dependency feature table of reference sentence 1 are successfully matched, then the total matching degree between the dependency feature table corresponding to the text information and the reference dependency feature table corresponding to the reference text information is changed from the initial value " 0” is updated to “20%”; then, the dependency feature table of statement 2 is compared with the reference dependency features corresponding to reference statement 1, reference statement 2 and reference statement 3 respectively, and the dependency relationship feature of statement 2 is obtained. The table is successfully matched with the reference dependency feature table of reference sentence 2, then the total matching degree between the dependency feature table corresponding to the text information and the reference dependency feature table corresponding to the reference text information is updated from "20%" to "50%" %"; finally, compare the dependency feature table of statement 3 with the reference dependency features corresponding to reference statement 1, reference statement 2 and reference statement 3, respectively, and obtain the dependency feature table of statement 3 and all reference statements. If the reference dependency feature table matching fails, the above total matching degree remains unchanged. Since all sentences in the text information have been compared, the comparison is stopped, and the above total matching degree is finally obtained as "50%".

Table 2

参考语句reference statement	参考匹配度reference match

参考语句1Reference statement 1	20％20%

参考语句2 Reference statement 2	30％30%
参考语句3Reference statement 3	50％50%

In some embodiments, all dependency feature sub-tables and reference dependency feature sub-tables may be obtained after sequentially comparing the dependency feature sub-tables of all statements in the text information with the reference dependency feature sub-tables of reference sentences. All comparison results of , any comparison result is used to indicate whether the corresponding dependency feature and the corresponding reference dependency feature sub-table are successfully matched. All comparison results can be parsed. If any comparison result indicates that the corresponding dependency feature is successfully matched with the corresponding reference dependency feature subtable, the total matching degree between the dependency feature table and the reference dependency feature table ( The initial value of the total matching degree is 0) by accumulating the reference matching degree corresponding to the reference dependency feature sub-table until all comparison results are analyzed, and the above-mentioned total matching degree can be obtained. In the above manner, the reference matching degree is allocated correspondingly according to the importance of each reference sentence, which can further improve the accuracy of the obtained total matching degree.

Wherein, each dependency feature sub-table and reference dependency feature sub-table includes the same N (N is an integer greater than 0) feature items, for example, the relationship dependency feature sub-table shown in FIG. 8 includes 7 feature items , respectively: predicate prototype, predicate lexical, subclass frame, path, location, dependency, and center word. The data corresponding to each feature item in any dependency feature subtable can be weighted and compared with the data corresponding to each feature item in any reference dependency feature subtable, and the dependency feature subtable can be determined based on the weighted comparison result. Whether it is successfully matched with any of the reference dependency feature sub-tables. Since the same feature items have comparative value, the above method for data comparison for the same feature items can ensure the validity of the data comparison.

Exemplarily, take any reference dependency feature sub-table as reference sub-table 1, and any dependency feature sub-table as sub-table 1 as an example, assuming that reference sub-table 1 and sub-table 1 both include N feature items, each A feature item can be preset with a corresponding weighted score K (K is a value greater than 0). In this case, the data corresponding to the N feature items in the sub-table 1 can be compared with the data corresponding to the N feature items in the reference sub-table in turn (compare the data corresponding to the same feature items), for example, the sub-table 1 Compare the data corresponding to the feature item "predicate prototype" in the reference subtable 1 with the data corresponding to the feature item "predicate prototype" in the reference subtable 1, and compare the data corresponding to the feature item "predicate lexical nature" in the subtable 1 with the feature item in the reference subtable 1. The data corresponding to the "predicate" is compared. In the comparison process, if the data corresponding to any feature item in the sub-table 1 is the same as the data corresponding to any one of the feature items in the reference sub-table, the approximate score between the sub-table 1 and the reference sub-table 1 can be calculated ( The initial value of the approximation score is 0) Accumulate the weighted score K corresponding to any feature item, and so on. When all the data corresponding to the N feature items are compared, the latest approximation of sub-table 1 and reference sub-table 1 can be obtained. Fraction. If the approximate score is greater than or equal to the approximate score threshold, it can be determined that the sub-table 1 and the reference sub-table 1 are successfully matched; otherwise, the matching fails.

Wherein, in the process of comparing the data corresponding to the N feature items, synonym replacement may also be considered for some specific feature items, and the specific feature items may include: predicate prototypes, central words, and the like. For example, the data corresponding to the feature item "predicate prototype" in sub-table 1 is: reinforcement; refer to the data corresponding to the feature item "predicate prototype" in sub-table 1: reinforcement; then, if the replacement of synonyms is not considered, compare the feature items " When the data corresponding to the "predicate prototype" is different, since the corresponding data of the feature item "predicate prototype" in the sub-table 1 and the reference sub-table 1 are different, it can be directly determined that the sub-table 1 and the reference sub-table 1 have different data in the feature item "predicate prototype". same. However, if the replacement of synonyms is considered, when it is determined that the corresponding data of the feature item "predicate prototype" in the judgment sub-table 1 and the reference sub-table 1 are different, it can also be judged whether the "enhancement" in the sub-table 1 is the "enhancement" in the reference sub-table. ”, if it is a synonym, then it can be determined that sub-table 1 and reference sub-table 1 have the same corresponding data in the feature item “predicate prototype”.

In the embodiment of the present application, the answer information and object information of the answering subjects who have completed the answering can be displayed on the answering progress page, which is beneficial to assist the subject who gave the question to correct the test paper and improve the efficiency of correcting the test paper.

Embodiments of the present application provide a computer program product or computer program, where the computer program product or computer program includes program instructions, and the program instructions are stored in a computer-readable storage medium. The processor of the electronic device reads the program instructions from the computer-readable storage medium, and the processor executes the program instructions, so that the electronic device executes the data processing methods described above in the embodiments of the present application.

Embodiments of the present application further provide a computer storage medium, where program instructions are stored in the computer storage medium, and when the program instructions are executed, are used to implement the corresponding methods described in the foregoing embodiments.

Please refer to FIG. 9 again, which is a schematic structural diagram of a data processing apparatus according to an embodiment of the present application. The data processing apparatus according to an embodiment of the present application may be set in the above-mentioned question generating terminal, or may be a computer running in the question generating terminal. Program (including program code).

In an implementation manner of the data processing apparatus of the embodiment of the present application, the data processing apparatus includes the following structure.

The display unit 90 is configured to display the test paper creation page;

The processing unit 91 is configured to trigger the acquisition of test paper data through the test paper creation page, and the test paper data is data that is triggered to upload or input through the test paper creation page;

The processing unit 91 is further configured to display the online test paper document according to the test paper data;

The publishing unit 92 is configured to publish the online test paper document.

In some embodiments, the processing unit 91 is further configured to, in response to an editing operation input on the online test paper document, perform editing processing on the online test paper document, and the editing processing is used to edit at least one of the following: the test questions of the target test questions in the online test paper document Content, question scores and reference answers.

In some embodiments, the online test paper document includes the test question content and answering area of each test question, and the processing unit 91 is further configured to display the answer editing entry of the target question in response to the triggering operation for the answering area corresponding to the target question; through the answer editing entry Trigger the display unit 90 to display the answer editing area; in response to inputting target information corresponding to the target test question in the answer editing area, display at least part of the target information in the online test paper document through the display unit 90; wherein the target information includes at least one of the following : The reference answer and question score of the target question.

In some embodiments, the online test paper document is displayed in the test paper editing page, and the test paper editing page is further displayed with a release entry, and the release unit 92 is further configured to: in response to a triggering operation for the release entry, generate an association with the online examination paper document The link of the test paper; share the link of the test paper to the answering terminal, and the link of the test paper is used: the answering terminal triggers the acquisition of the target test paper document that matches the online test paper document through the test paper link, so that the respondents can answer the target test paper document through the answering terminal.

In some embodiments, the publishing unit 92 is further configured to: display the test paper link and the sharing button through the display unit 90, trigger the determination of the target sharing method and the target sharing address through the sharing button, and share the test paper link according to the target sharing method and the target sharing address to the answering terminal.

In some embodiments, the display unit 90 is further configured to display an answering progress page associated with the online test paper document, and the answering progress page includes answering information and object information of the answering subjects who have completed the answering; The viewing operation of the target answering object displays the detailed information of the test paper of the target answering object. The detailed information of the test paper includes at least one of the following: the reference answer of each test question in the online test paper document, the answer content entered by the target answering object when answering each test question, and the score As a result, the scoring result includes the predicted answer score obtained by the target answering object in each test question.

In some embodiments, each test question includes a subjective question and an objective question, and the processing unit 91 is further configured to: obtain the target answer content entered by the target answering object when answering the subjective question, and the target reference answer of the subjective question; When the reference formula is included, formula recognition is performed on the formula in the target answer content to obtain the formula recognition result, and the predicted answer score obtained by the target answer subject in the subjective question is determined according to the formula recognition result; when the target reference answer includes reference text information, the target The text information in the answer content is semantically identified to obtain the semantic identification result, and the answer prediction score obtained by the target answer object in the subjective question is determined according to the semantic identification result.

In some embodiments, the processing unit 91 is further configured to: perform word segmentation processing on the reference text information in the target reference answer to obtain at least one reference word segment; Keyword matching is performed on the textual information of the textual information, and the semantic recognition result is obtained; when any reference participle in at least one reference participle is not a same-sex keyword, and the number of historical reference scoring data existing in the subjective question meets the quantitative condition, the content of the target answer is determined. The text information in the text information is judged and scored, and the semantic recognition result is obtained; when any reference participle in at least one reference participle is not a same-sex keyword, and the number of historical reference scoring data existing in the subjective question does not meet the quantitative condition, the target The text information in the answer content is analyzed by dependency syntax, and the semantic recognition result is obtained.

In some embodiments, the processing unit 91 is further configured to: perform word segmentation on the text information in the target answer content to obtain at least one word segment; match at least one word segment and at least one reference word segment, and determine the matching result as a semantic recognition result.

In some embodiments, the processing unit 91 is further configured to: invoke the target scoring model to perform scoring processing on the text information in the target answer content to obtain a scoring result of the subjective question; determine the scoring result of the subjective question as semantic recognition As a result, the target scoring model is obtained by training the initial scoring model based on the historical reference scoring data of subjective questions.

In some embodiments, the processing unit 91 is further configured to: identify the sentences included in the text information in the target answer content, and perform a dependency relationship analysis on the sentences to determine the dependency feature table corresponding to the text information; obtain the reference text information corresponding to The reference dependency feature table is compared; the dependency feature table and the reference dependency feature table are compared, and the comparison result is determined as the semantic recognition result.

In some embodiments, when the test paper data is data triggered to be uploaded through the test paper creation page, the test paper data is parsed, and an online test paper document is generated based on the analysis result; the analysis result includes at least one of the following: the test questions of each test question in the test paper data Content, question stem, answer area, question type, reference answer, and target score information of test paper data.

In the embodiments of the present application, for the implementation of the foregoing units, reference may be made to the descriptions of the relevant contents in the embodiments corresponding to the foregoing respective drawings.

The data processing device in the embodiment of the present application can display the test paper creation page, trigger the acquisition of test paper data through the test paper creation page, and display the online test paper document according to the test paper data, thereby publishing the online test paper document. In the process of creating the test paper, there is no need to limit the type of test questions and the filling format of the test content, which is beneficial to greatly improve the flexibility of test paper creation.

Please refer to FIG. 10 again, which is a schematic structural diagram of a terminal according to an embodiment of the present application. The terminal here may refer to the above-mentioned terminal for generating questions, including but not limited to: tablet computer, laptop computer, notebook computer, and desktop computer, etc. . The terminal in this embodiment of the present application includes a power supply module and other structures, and includes a processor 100, a storage device 101, an input device 102, an output device 103, and a communication interface 104. Data can be exchanged among the processor 100 , the storage device 101 , the input device 102 , the output device 103 and the communication interface 104 , and the processor 100 implements corresponding data processing functions.

The storage device 101 may include a volatile memory (Volatile Memory), such as a random access memory (Random-Access Memory, RAM); the storage device 101 may also include a non-volatile memory (Non-Volatile Memory), such as a flash memory (Flash Memory), Solid-State Drive (Solid-State Drive, SSD), etc.; the storage device 101 may also include a combination of the above-mentioned types of memories.

The processor 100 may be a central processing unit 100 (Central Processing Unit, CPU). In some embodiments, the processor 100 may also be a graphics processor 100 (Graphics Processing Unit, GPU). The processor 100 may also be a combination of a CPU and a GPU. In the terminal, multiple CPUs and GPUs can be included to perform corresponding data processing as required.

The input device 102 can refer to a display screen, a fingerprint collector, etc., and can be used to detect related operations input by the user (such as editing operations for online test paper document input, triggering operations for the answering area corresponding to the target test question, etc.), and the output device 103 can be Including displays (LCD, etc.), speakers, etc.

In some embodiments, storage device 101 is used to store program instructions. The processor 100 may invoke program instructions to implement various methods as mentioned above in the embodiments of the present application.

In some embodiments, the processor 100 of the terminal invokes the program instructions stored in the storage device 101 to: display the test paper creation page through the output device 103; trigger the acquisition of test paper data through the test paper creation page, and the test paper data is the test paper creation page Trigger the uploaded or input data; display the online test paper document according to the test paper data; publish the online test paper document through the communication interface 104 .

In this embodiment of the present application, for the implementation of the above-mentioned processor 100, reference may be made to the description of the relevant content in the embodiments corresponding to the foregoing figures.

The terminal in the embodiment of the present application may display a test paper creation page, trigger the acquisition of test paper data through the test paper creation page, and display the online test paper document according to the test paper data, thereby publishing the online test paper document. In the process of creating the test paper, there is no need to limit the type of test questions and the filling format of the test content, which is beneficial to greatly improve the flexibility of test paper creation.

Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be implemented by instructing the relevant hardware through a computer program, and the described program can be stored in a computer-readable storage medium. During execution, the processes of the embodiments of the above-mentioned methods may be included. The computer-readable storage medium may be a magnetic disk, an optical disk, a read-only memory (Read-Only Memory, ROM), or a random access memory (Random Access Memory, RAM), or the like.

The above disclosure is only a part of the embodiments of the present application, of course, the scope of the rights of the present application cannot be limited by this. Those of ordinary skill in the art can understand that all or part of the procedures for realizing the above-mentioned embodiments are implemented according to the claims of the present application. The equivalent changes are still within the scope of the application.

Claims

A data processing method, executed by an electronic device, the method comprising:

Display the test paper creation page;

Triggering the acquisition of test paper data through the test paper creation page, where the test paper data is data that is triggered to upload or input through the test paper creation page;

Display the online test paper document according to the test paper data;

Publish the online test paper documentation.
The method of claim 1, wherein after the online test paper document is displayed according to the test paper data, the method further comprises:

In response to an editing operation input for the online test paper document, an editing process is performed on the online test paper document, and the editing process is used to edit at least one of the following: test question content, test question score value of the target test question in the online test paper document and reference answers.
The method according to claim 2, wherein the online test paper document includes test question content and answering area of each test question; the online test paper document is edited in response to an input editing operation for the online test paper document ,include:

In response to the trigger operation corresponding to the answering area of the target test question, the answer editing entry of the target test question is displayed;

Displaying an answer editing area through the answer editing entry;

In response to inputting target information corresponding to the target question in the answer editing area, displaying at least part of the target information in the online test paper document;

Wherein, the target information includes at least one of the following: the reference answer of the target test question and the test question score.
The method according to any one of claims 1 to 3, wherein the online test paper document is displayed on a test paper editing page, and the test paper editing page is further displayed with a publishing entry; the publishing the online test paper document, include:

in response to a triggering operation for the release portal, generating a test paper link associated with the online test paper document;

The test paper link is shared to the answering terminal, and the test paper link is used for: the answering terminal obtains the target test paper document that matches the online test paper document by triggering the test paper link, so that the answering object can pass the answering terminal. The target test paper document is used for answering questions.
The method of claim 4, wherein the sharing of the test paper link to the answering terminal comprises:

display the test paper link and share button;

Triggering and determining the target sharing method and target sharing address through the sharing button;

The test paper link is shared to the answering terminal according to the target sharing method and the target sharing address.
The method according to claim 5, wherein the triggering and determining of the target sharing method and target sharing address through the sharing button comprises:

In response to the trigger operation for the sharing button, displaying a sharing method selection list;

In response to the selection operation for multiple sharing methods in the sharing method selection list, the selected sharing method is used as the target sharing method;

The target sharing address is determined according to the target sharing method.
The method of claim 4, wherein after the test paper link is shared to the answering terminal, the method further comprises:

Displaying the answering progress page associated with the online test paper document, the answering progress page including the answering information and object information of the answering subjects who have completed the answering;

In response to the viewing operation for the target answering object in the answering progress page, the detailed information of the test paper of the target answering object is displayed, and the detailed information of the test paper includes at least one of the following: the reference answer of each test question in the online test paper document , the answer content input by the target answering object when answering each test question, and the scoring result, the scoring result including the answering prediction score obtained by the target answering object in each test question.
The method of claim 7, wherein each test question includes a subjective question and an objective question; the method further comprises:

Obtaining the target answer content entered by the target answering object when answering the subjective question, and the target reference answer of the subjective question;

When the target reference answer includes a reference formula, formula recognition is performed on the formula in the target answer content to obtain a formula recognition result, and the answer prediction obtained by the target answer object in the subjective question is determined according to the formula recognition result. Score;

When the target reference answer includes reference text information, semantic recognition is performed on the text information in the target answer content to obtain a semantic recognition result, and according to the semantic recognition result, it is determined that the target answer object obtained in the subjective question Answer prediction score.
The method according to claim 8, wherein the semantic recognition result obtained by performing semantic recognition on the text information in the target answer content comprises:

Perform word segmentation processing on the reference text information in the target reference answer to obtain at least one reference word segmentation;

When the at least one reference word segment is a same-sex keyword, keyword matching is performed on the text information in the target answer content to obtain a semantic recognition result;

When any reference participle in the at least one reference participle is not a same-sex keyword, and the number of historical reference scoring data existing in the subjective question satisfies the quantitative condition, the text information in the target answer content is judged processing to obtain semantic recognition results;

When any reference participle in the at least one reference participle is not a same-sex keyword, and the quantity of historical reference scoring data existing in the subjective question does not satisfy the quantity condition, the text information in the target answer content Dependency syntax analysis is performed to obtain semantic recognition results.
The method according to claim 9, wherein, performing keyword matching on the text information in the target answer content to obtain a semantic recognition result, comprising:

Perform word segmentation processing on the text information in the target answer content to obtain at least one word segmentation;

The at least one participle and the at least one reference participle are matched, and the matching result is determined as a semantic recognition result.
The method of claim 10, wherein the matching the at least one participle and the at least one reference participle comprises:

Perform any of the following processing:

processing the at least one word segmentation into a keyword sequence, processing the at least one reference word segmentation into a reference keyword sequence, and matching the keyword sequence with the reference keyword sequence;

Perform unordered matching on the at least one participle and the at least one reference participle.
The method according to claim 9, wherein, performing a scoring process on the text information in the target answer content to obtain a semantic recognition result, comprising:

Invoke the target scoring model to perform scoring processing on the text information in the target answer content, and obtain a subjective question scoring result;

Determining the subjective question scoring result as the semantic recognition result;

Wherein, the target scoring model is obtained by training the initial scoring model based on the historical reference scoring data of the subjective question.
The method according to claim 9, wherein, performing a dependency syntax analysis on the text information in the target answer content to obtain a semantic recognition result, comprising:

Identifying the statement included in the text information in the target answer content, and performing a dependency relationship analysis on the statement to obtain a dependency feature table corresponding to the text information;

obtaining a reference dependency feature table corresponding to the reference text information;

Comparing the dependency feature table with the reference dependency feature table, a semantic recognition result is obtained.
The method according to claim 13, wherein, performing a dependency relationship analysis on the statement to obtain a dependency relationship feature table corresponding to the text information, comprising:

For any statement included in the text information, perform the following processing:

Performing a dependency relationship analysis on any of the sentences to obtain the part of speech, role classification and dependencies of each participle in the any sentence;

According to the part-of-speech, role classification and dependency of each participle in any of the sentences, a dependency tree corresponding to the any of the sentences is generated;

According to the dependency tree corresponding to any statement, determine the dependency feature sub-table of any statement;

Wherein, the dependency feature sub-table of each statement in the text information is used to construct a dependency feature table corresponding to the text information.
The method according to claim 13, wherein the dependency relationship feature table includes a dependency relationship feature sub-table of each statement in the text information, and the reference dependency relationship feature table includes a dependency relationship feature table of each reference statement in the reference text information. Refer to the dependency feature subtable;

Said comparing the dependency feature table and the reference dependency feature table to obtain a semantic recognition result, including:

comparing the dependency feature sub-table of each statement in the text information with the reference dependency feature sub-table of each reference statement in the reference text information;

Based on the comparison result between any dependency feature sub-table and any reference dependency feature sub-table, determine the matching degree between the any dependency feature sub-table and any reference dependency feature sub-table;

A total matching degree is obtained by accumulating a plurality of the matching degrees, and the total matching degree is used as a semantic recognition result.
The method of claim 15, wherein each reference sentence in the reference text information is assigned a reference matching degree; the comparison result based on any dependency feature sub-table and any reference dependency feature sub-table, Determining the degree of matching between any of the dependency feature sub-tables and any of the reference dependency feature sub-tables, including:

When the comparison result between the any dependency feature sub-table and the any reference dependency feature sub-table indicates that the matching is successful, the reference matching degree assigned to the reference sentence corresponding to the any reference dependency feature sub-table , as the matching degree between any of the dependency feature sub-tables and any reference dependency feature sub-table;

When the comparison result between the any dependency feature sub-table and the any reference dependency feature sub-table indicates that the matching fails, determine that the any dependency feature sub-table and the any reference dependency feature sub-table The match between them is zero.
The method according to claim 15, wherein the comparing the dependency feature sub-table of each statement in the text information with the reference dependency feature sub-table of each reference statement in the reference text information comprises:

The data corresponding to each feature item in any dependency feature subtable is weighted and compared with the data corresponding to each feature item in any reference dependency feature subtable to obtain the any dependency feature subtable and the any one Refer to the comparison results of the dependency feature sub-table.
The method according to any one of claims 1 to 3, wherein after triggering the acquisition of test paper data through the test paper creation page, the method further comprises:

When the test paper data is data triggered to be uploaded through the test paper creation page, the test paper data is parsed, and an online test paper document is generated based on the analysis result; the analysis result includes at least one of the following: in the test paper data Test question content, question stem, answering area, test question type, reference answer, and target score information of the test paper data of each test question.
The method of claim 18, wherein the analyzing the test paper data, and generating an online test paper document based on the analysis result, comprises:

Identify the first-level serial number in the test paper data and the text content corresponding to the first-level serial number;

Match the text content corresponding to the first-level serial number with the keyword corresponding to each test question type, and determine the test question type of each test question in the test paper data according to the matching result;

Determine the question stem and answer area of each test question according to the test question type of each test question;

According to the test question type, question stem and answering area of each test question, the test content of each test question is integrated and processed to obtain an online test paper document.
The method according to claim 19, wherein after determining the test question type of each test question in the test paper data according to the matching result, the method further comprises:

Parse the reference answer of each test question at the target position associated with the test question type of each test question;

Obtain the score-related information corresponding to each test question type in the test paper data, and analyze the score-related information to obtain the target score information of the test paper data;

Wherein, the reference answer of each test question and the target score information of the test paper data are used to combine the test question type, question stem, answering area and test question content of each test question to generate an online test paper document.
The method according to any one of claims 1 to 3, wherein the test paper creation page includes a first entry and a second entry; and the triggering acquisition of test paper data through the test paper creation page includes:

Perform any of the following processing:

In response to the trigger operation for the first entry, obtain the uploaded test paper data;

In response to the triggering operation for the second entry, an online document is created, and the data entered in the online document is used as test paper data.
The method according to any one of claims 1 to 3, wherein the online test paper document is displayed on a test paper editing page, and the test paper editing page includes an answer configuration entry; the method further comprises:

in response to a triggering operation for the answer configuration entry, displaying an answer configuration window;

The answering configuration information is obtained through the answering configuration window, and answering configuration is performed on the online test paper document according to the answering configuration information.
A data processing device, comprising:

A display unit, configured to display the test paper creation page;

a processing unit, configured to trigger the acquisition of test paper data through the test paper creation page, where the test paper data is data that is triggered to upload or input through the test paper creation page;

The processing unit is further configured to display the online test paper document according to the test paper data;

A publishing unit configured to publish the online test paper document.
An electronic device, the electronic device includes a processor and a storage device, the processor and the storage device are connected to each other, wherein the storage device is used to store a computer program, the computer program includes program instructions, the A processor is configured to invoke the program instructions to perform the method of any of claims 1-22.
A computer storage medium, where program instructions are stored in the computer storage medium, and when the program instructions are executed, are used to implement the method according to any one of claims 1-22.
A computer program product comprising program instructions which, when executed by a processor, implement the method of any one of claims 1-22.