Intelligent supervision method for handwriting answering of subjective questions in learning and machine-based examination

Abstract

The invention discloses an intelligent supervision method for handwriting answering of subjective questions in machine-based examination of learning examination, and belongs to the technical field of machine-based examination of subjective questions, the method comprises the following steps: step 1, setting and updating a handwriting device directory in real time, the handwriting device directory being used for counting information of each handwriting device; the method comprises the following steps: setting a modular acquisition module according to a handwriting device directory, wherein the modular acquisition module consists of corresponding modular units; 2, acquiring machine-based examination information, recommending a handwriting device combination to a manager according to the machine-based examination information, and determining a corresponding handwriting application device by the manager; 3, connecting the modularized acquisition module with corresponding handwriting application equipment to form an equipment pool; carrying out real-time data acquisition according to the equipment pool to obtain handwritten acquisition data; 4, processing the handwritten collected data to obtain answer data; the monitoring result and the answering data are displayed to the examinee; and the examinee performs corresponding adjustment according to the answer data.

Description

Technical Field

[0001] The present invention belongs to the technical field of subjective question computerized examinations, and specifically relates to an intelligent supervision method for handwritten answers to subjective questions in academic examinations using a computerized examination system. Background Art

[0002] In the traditional paper-and-pencil examination environment, subjective questions such as short-answer questions, analysis questions, and essays have always occupied an important position. These question types can comprehensively examine students' knowledge understanding, analysis ability, and innovative thinking. However, with the rapid development of information technology, more and more examinations are turning to computerized examination forms in order to improve the efficiency, fairness, and security of examinations. However, how subjective questions adapt to this transformation has become an urgent problem to be solved.

[0003] Currently, for computerized examinations of subjective questions, various solutions have emerged, but each has obvious limitations. For example, attempts have been made to convert subjective questions into objective questions for computerized examinations, such as indirectly examining students' subjective abilities through multiple-choice questions, true-or-false questions, etc. However, this conversion often fails to accurately measure students' true levels and abilities because the core of subjective questions lies in examining students' analysis, reasoning, and expression abilities, and these abilities are difficult to comprehensively evaluate through simple objective questions. Another common solution is to allow students to answer questions on the examination system using a keyboard or a stylus. Although keyboard answering is convenient, it is limited to text input and is clearly inadequate for subjects that require graphics, formulas, or complex mathematical calculations (such as mathematics, physics, chemistry, etc.). There is also handwritten answering using a dot matrix pen, but dot matrix pen answering requires special answer sheets, and the answer sheets need to be bound to the test questions before the exam. This not only increases the preparation workload of the examination but also may lead to examination errors due to improper binding of the answer sheet to the test questions. Therefore, the existing various solutions have certain usage defects.

[0004] Based on this, in order to solve the limitations of computerized examinations of subjective questions, the present invention provides an intelligent supervision method for handwritten answers to subjective questions in academic examinations using a computerized examination system. Summary of the Invention

[0005] In order to solve the problems existing in the above solutions, the present invention provides an intelligent supervision method for handwritten answers to subjective questions in academic examinations using a computerized examination system.

[0006] The object of the present invention can be achieved by the following technical solutions: An intelligent supervision method for handwritten answers to subjective questions in academic examinations using a computerized examination system, the method comprising: Step 1: Set and update in real time a handwritten device directory, where the handwritten device directory is used to count information of each handwritten device; set a modular acquisition module according to the handwritten device directory, and the modular acquisition module is composed of corresponding modular units, and the modular units are used to connect to and collect data from corresponding handwritten devices; Furthermore, the method for setting and updating the handwriting device list in real time includes: Setting a handwriting device directory template, and setting corresponding device feature items according to the handwriting device directory template; Collecting each computer-based test announcement data, extracting features from the computer-based test announcement data according to the device feature items, and obtaining the candidate's writing device information; deduplicating the candidate's writing device information, marking the remaining candidate's writing device information as handwriting device information, and adding the handwriting device information to the handwriting device list template to obtain a handwriting device list; The handwriting device information to be selected is collected in real time, the handwriting device information to be selected is screened according to the handwriting device list, the handwriting device information is determined, and the handwriting device information is added to the handwriting device list.

[0007] Furthermore, the method for screening the candidate's handwriting device information according to the handwriting device list includes: Marking each piece of handwriting device information in the handwriting device list as a set element; performing real-time recognition on the handwriting device list to obtain the set element corresponding to the handwriting device list, and forming a verification set of the handwriting device list from the set elements; Establish a screening model, the expression of the screening model is: ; Where: (s, U) is the input data, s is the device information to be written, and U is the verification set; s∈U means that the device information to be written belongs to the verification set; the output data is the verification screening value SU(s, U), and the verification screening value is 1 or 0; Analyze the verification set and the device information to be selected by using the screening model to obtain the verification screening value of the device information to be selected; When the check screening value is 1, the candidate writing device information is eliminated; When the verification screening value is 0, the handwriting device information to be selected is marked as handwriting device information.

[0008] Step 2: Obtain computer-based test information, recommend a handwriting device combination to the management personnel based on the computer-based test information, and the management personnel determine the corresponding handwriting application device; Further, the method for recommending a handwriting device combination to a manager based on the computer-based test information includes: A preset handwriting feature table is used to count corresponding handwriting features; Identify computer-based test information, perform feature analysis on the computer-based test information according to the handwriting feature table, obtain handwriting features corresponding to the computer-based test information, and integrate the handwriting features into computer-based test equipment requirements for the computer-based test information; Obtain a list of handwriting devices, analyze the list of handwriting devices according to the requirements of the computerized examination devices, and obtain a combination of candidate devices; Screen the combination of candidate devices to determine the recommended combination of handwriting devices.

[0009] Further, the method for screening the combination of candidate devices includes: Obtain the computerized examination conditions, and estimate the additional cost of the combination of candidate devices according to the computerized examination conditions; Obtain a handwriting feature table, set the handwriting values of the corresponding handwriting devices for the handwriting features in the handwriting feature table; establish a handwriting value statistical table according to the handwriting features, handwriting devices, and handwriting values; Identify each handwriting device corresponding to the combination of candidate devices, and mark it as a unit device; determine the handwriting features that the unit devices in the combination of candidate devices need to solve according to the requirements of the computerized examination devices, and match the handwriting values of the corresponding handwriting features from the handwriting value statistical table according to the unit devices and handwriting features; Mark the handwriting value as SY i , where i represents the corresponding handwriting feature in the requirements of the computerized examination devices, i = 1, 2,..., n, and n is the number of handwriting features in the requirements of the computerized examination devices; Calculate the screening value of the corresponding combination of candidate devices according to the screening formula, and determine the recommended combination of handwriting devices according to the screening value.

[0010] Further, the screening formula is: ; In the formula: SA is the screening value; b 1 , b 2 are both proportionality coefficients, and the value range is 0 < b 1 ≤1, 0 < b 2 ≤1; XB is the additional cost.

[0011] Step 3: Connect the modular acquisition module with the corresponding handwriting application device to form a device pool; perform real-time data acquisition according to the device pool to obtain handwriting acquisition data; Step 4: Process the handwriting acquisition data to obtain answer data; display the monitoring results and the answer data to the examinee; the examinee makes corresponding adjustments according to the answer data.

[0012] Further, perform anomaly monitoring on the answer data to obtain monitoring results; supplement the monitoring results to the answer data.

[0013] Further, the method for performing anomaly monitoring on the answer data includes: Set anomaly monitoring items and the corresponding anomaly criteria; establish an anomaly recognition model according to the anomaly monitoring items and the anomaly criteria; Collect data from the answer data according to the abnormal monitoring items to obtain the monitoring item data corresponding to the abnormal monitoring items; Analyze the monitoring item data through the abnormal recognition model to obtain the corresponding monitoring results.

[0014] Compared with the prior art, the beneficial effects of the present invention are as follows: Through the present invention, the integration of multiple input methods such as keyboards, writing tablets, and dot matrix pens is realized, solving the problems of the use defects of existing computerized examinations; it can present keyboard input content, handwriting input content, dot matrix pen input content, etc. in the same answer area, realizing functions that cannot be achieved by all existing rich text editors; it provides a more convenient answering experience for candidates, and candidates can freely switch inputs according to the needs of the answering content; the answering process of candidates can be displayed through playback. Brief Description of the Drawings

[0015] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0016] Figure 1 It is a flowchart of the method of the present invention. Detailed Embodiments

[0017] The following will clearly and completely describe the technical solutions of the present invention in conjunction with the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the protection scope of the present invention.

[0018] As Figure 1 shown, a smart supervision method for handwritten answers to subjective questions in academic examinations by computerized examinations includes: Step 1: Set and update the handwritten device list in real time. The handwritten device list is used to count the information of each handwritten device that meets the requirements of handwritten computerized examinations, such as mice, keyboards, writing tablets, dot matrix pens, etc.; set a modular acquisition module according to the handwritten device list. The modular acquisition module is used to connect the handwritten devices of the examination application and perform corresponding data acquisition. The modular acquisition module is composed of each modular unit, and the modular unit is used to connect and collect data from the handwritten devices within the acquisition device range; subsequently, judge whether it is necessary to update and set new modular units or update the corresponding acquisition device range according to the update of the handwritten device list.

[0019] In one embodiment, the handwritten device directory can be set and updated based on existing methods, such as setting and updating it manually.

[0020] In one embodiment, the method for setting and real-time updating the handwritten device directory includes: Set a handwritten device directory template, which is used to represent what information of the handwritten device needs to be counted, such as relevant information like name, model, communication protocol, etc.; Set corresponding device feature items according to the handwritten device directory template, that is, subsequently collect corresponding handwritten device information according to the feature items, such as device feature items corresponding to name, model, etc.; Collect each computer-based examination announcement data, which needs to include handwritten device data that meets the requirements or is allowed to be used. Extract features from the collected computer-based examination announcement data according to the device feature items to obtain corresponding candidate handwritten device information; remove duplicates from the obtained candidate handwritten device information, mark the remaining candidate handwritten device information as handwritten device information, and supplement the handwritten device information into the handwritten device directory template to obtain the handwritten device directory; Real-time collect candidate handwritten device information, screen the corresponding candidate handwritten device information according to the handwritten device directory, determine the handwritten device information, and supplement the handwritten device information into the handwritten device directory.

[0021] The various handwritten device information in the handwritten device directory can be classified based on the current classification method.

[0022] In one embodiment, the method for screening the corresponding candidate handwritten device information according to the handwritten device directory includes: Mark each handwritten device information in the handwritten device directory as a set element; perform real-time recognition on the handwritten device directory to obtain each set element corresponding to the handwritten device directory, and form a verification set corresponding to the handwritten device directory from each set element; Establish a screening model, and the expression of the screening model is: ; In the formula: (s, U) is the input data, s is the candidate handwritten device information, and U is the verification set; s ∈ U means that there is a set element in the verification set that is the same as the candidate handwritten device information, that is, the candidate handwritten device information belongs to the verification set; the output data is the verification screening value SU(s, U), and the verification screening value is 1 or 0; Analyze the verification set and the candidate handwritten device information through the screening model to obtain the verification screening value of the corresponding candidate handwritten device information; When the verification screening value is 1, eliminate the candidate handwritten device information; When the verification screening value is 0, mark the candidate handwritten device information as handwritten device information.

[0023] In one embodiment, the modular acquisition module can be set according to the list of handwriting devices and established based on existing methods.

[0024] Exemplarily, identify the handwriting device information in the list of handwriting devices, determine the corresponding connection acquisition method according to the handwriting device information, and determine it according to what is allowed; classify the corresponding handwriting device information according to the connection acquisition method to obtain the corresponding acquisition classification, that is, those that can be connected and acquired using the same modular acquisition unit after analysis are classified into one category. Set the corresponding modular acquisition unit according to each acquisition classification. The modular acquisition unit can be directly classified and set manually, and the acquisition device range of the modular acquisition unit is formed according to the acquisition classification. The acquisition device range may exceed the corresponding acquisition classification according to the actually set modular acquisition unit; set the modular acquisition module according to each modular acquisition unit; Subsequently, the modular acquisition module is updated accordingly according to the update of the list of handwriting devices.

[0025] Step 2: Obtain computer-based examination information. The computer-based examination information includes information such as examination subjects and examination scopes. According to the computer-based examination information, recommend a combination of handwriting devices to the management personnel. The management personnel determine each handwriting device to be applied and mark it as a handwriting application device.

[0026] In one embodiment, the method for recommending a combination of handwriting devices to the management personnel according to the computer-based examination information includes: Preset a handwriting feature table, which is used to count various handwriting features that will be encountered in the examination. It is mainly set for the subjective questions of each subject, such as text features, mathematical formula features, chemical formula features, etc. Specifically, it is set by the platform party according to the handwriting differences in computer-based examinations, and can be summarized by combining historical computer-based examination data, examination records, etc., and then screened and verified manually to form a handwriting feature table; Identify the computer-based examination information, perform feature analysis on the computer-based examination information according to the handwriting feature table, and determine the possible handwriting features of the computer-based examination information. Since it is an analysis of the examination scope, there are multiple possibilities. Avoid directly identifying and analyzing examination answers, etc., which may affect the examination effect, such as calculating whether there are certain formulas, leakage of examination questions, etc.; integrate each handwriting feature into the computer-based examination device requirements; Analyze the list of handwriting devices according to the computer-based examination device requirements, and determine the combination of handwriting devices that can meet the computer-based examination device requirements, which is marked as the candidate device combination. For example, if the combination of a keyboard and a handwriting tablet can meet each handwriting feature in the computer-based examination device requirements, then its combination is the candidate device combination. Specifically, the combination analysis is performed according to the main handwriting features solved by each handwriting device. For example, the keyboard is mainly used to solve handwriting features such as text and Chinese symbols; Screen the candidate device combinations to determine the recommended handwriting device combinations.

[0027] In one embodiment, screening the candidate device combinations can be performed based on existing screening algorithms, such as screening from the perspective of cost.

[0028] In one embodiment, the method for screening candidate device combinations includes: Obtain the existing computer-based examination conditions, that is, various existing handwriting devices, such as keyboards and mice; estimate the additional cost for each candidate device combination according to the computer-based examination conditions, such as the additional cost of adding a handwriting device that the current organizer does not have, and obtain the additional cost of the corresponding candidate device combination; Obtain the handwriting feature table, and collect the handwriting efficiency of each handwriting feature in the handwriting feature table using each handwriting device, that is, calculate according to the time required to use different handwriting devices for the same answer. Through statistics on a large amount of historical data, form a representative handwriting efficiency of each handwriting device for the corresponding handwriting feature, marked as the handwriting value, and the average value, mode, etc. can be selected for determination; establish a handwriting value statistical table according to the handwriting feature, handwriting device, and handwriting value; Identify each handwriting device corresponding to the candidate device combination, marked as a unit device; determine the handwriting features that each unit device in the candidate device combination needs to solve according to the computer-based examination device requirements, and match the handwriting value corresponding to the corresponding handwriting feature from the handwriting value statistical table according to the unit device and the handwriting feature; mark the obtained handwriting value as SY i , where i represents the corresponding handwriting feature in the computer-based examination device requirements, i = 1, 2,..., n, and n is the number of handwriting features in the computer-based examination device requirements; Calculate the screening value of the corresponding candidate device combination according to the screening formula. The screening formula is: ; In the formula: SA is the screening value; b 1 , b 2 are both proportionality coefficients, and the value range is 0 < b 1 ≤1, 0 < b 2 ≤1; XB is the additional cost; Select the candidate device combination with the largest screening value as the recommended handwriting device combination.

[0029] In other embodiments, the priorities of two candidate device combinations can also be directly compared comprehensively according to the ratio of the cumulative handwriting values and the proportion of the additional cost, that is: .

[0030] Step 3: Connect the modular acquisition module to the corresponding handwriting application device to form a device pool for counting the connected handwriting application devices; perform real-time data acquisition based on the device pool to obtain corresponding handwriting acquisition data, which is marked with corresponding tags such as devices and time.

[0031] Step 4: Process the handwriting acquisition data to obtain answer data. The answer data is marked with the corresponding handwriting application device marks for each part, which is convenient for candidates to check and adjust later; and perform anomaly monitoring on the answer data to obtain a monitoring result, and display the monitoring result and the answer data to the corresponding candidates; the candidates make corresponding adjustments according to the answer data.

[0032] In one embodiment, processing the handwriting acquisition data means using the existing method for processing, that is, processing in combination with the relevant technologies of the corresponding handwriting application device.

[0033] Exemplarily, encapsulate the handwriting acquisition data (add device number, timestamp, and checksum); splice the classified data packets to the existing data (save the continuous answer data); Paginate the data according to the page number in the data header (the dot matrix pen data has a page number in addition to the coordinates). If it is paginated, split the data into different pages; Perform coordinate conversion on the paginated data, convert the input coordinates into the x and y axis values of the image relative to the upper left corner, and finally store it as ordered data; Layer the received data according to the source and signal type (that is, convert it into N layers); For the data in each layer, use the SVG conversion numerical function to convert x / y into SVG objects; Call WebEngine to render the SVG objects using H5 technology to render N layer images.

[0034] In one embodiment, performing anomaly monitoring on the answer data refers to whether there are abnormal situations during the data processing process, which is used to prompt the user to avoid the user ignoring the abnormal situation. It mainly detects clarity, text fluency, sentence smoothness, etc. The anomaly monitoring items are specifically set by the administrator according to the examination situation.

[0035] Exemplarily, set each anomaly monitoring item and the corresponding anomaly standard, establish an anomaly recognition model according to each anomaly monitoring item, and the anomaly recognition model is used to analyze the monitoring item data corresponding to each anomaly monitoring item to determine whether it meets the corresponding anomaly standard, and then determine the monitoring result; The anomaly recognition model can be established based on a CNN network, a DNN network, etc., or based on existing recognition and judgment technologies, as long as it can achieve the recognition and judgment of whether the data of the corresponding monitoring item meets the anomaly standard.

[0036] In one embodiment, the anomaly monitoring of the answer data can also be anomaly monitoring such as cheating, and whether there are abnormal behaviors is analyzed according to the writing process, etc.

[0037] In one embodiment, the examinee makes corresponding adjustments according to the answer data, such as modifying text, formulas, adjusting the order, etc.

[0038] The above formulas are all calculated by removing the dimension and taking their numerical values. The formula is obtained by collecting a large amount of data for software simulation to get a formula closest to the actual situation. The preset parameters and preset thresholds in the formula are set by those skilled in the art according to the actual situation or obtained by simulating a large amount of data.

[0039] The above embodiments are only used to illustrate the technical method of the present invention and not to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical method of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical method of the present invention.

Claims

1. An intelligent supervision method for handwritten answers to subjective questions in computer-based exams, characterized in that: Methods include: Step 1: Setting and updating the handwriting device list in real time, wherein the handwriting device list is used to count the information of each handwriting device; According to the handwriting device list, a modular acquisition module is set, wherein the modular acquisition module is composed of corresponding modular units, and the modular units are used to connect with the corresponding handwriting devices and collect data; Step 2: Obtain computer-based test information, recommend a handwriting device combination to the management personnel based on the computer-based test information, and the management personnel determine the corresponding handwriting application device; Step 3: The modular acquisition module is connected with the corresponding handwriting application device to form a device pool; Perform real-time data collection according to the device pool to obtain handwriting collection data; Step 4: Process the handwriting collection data to obtain answer data; present the monitoring results and answer data to the examinee; Candidates make corresponding adjustments based on the answer data.

2. According to claim 1, a method for intelligent supervision of handwritten answers to subjective questions in a computer-based academic examination, characterized in that: The methods for setting and updating the handwriting device list in real time include: Setting a handwriting device directory template, and setting corresponding device feature items according to the handwriting device directory template; Collecting each computer-based test announcement data, extracting features from the computer-based test announcement data according to the device feature items, and obtaining the candidate's writing device information; deduplicating the candidate's writing device information, marking the remaining candidate's writing device information as handwriting device information, and adding the handwriting device information to the handwriting device list template to obtain a handwriting device list; The handwriting device information to be selected is collected in real time, the handwriting device information to be selected is screened according to the handwriting device list, the handwriting device information is determined, and the handwriting device information is added to the handwriting device list.

3. According to claim 2, a method for intelligent supervision of handwritten answers to subjective questions in a computer-based academic examination, characterized in that: Methods for screening the handwriting device information of candidates according to the handwriting device list include: Marking each piece of handwriting device information in the handwriting device list as a set element; performing real-time recognition on the handwriting device list to obtain the set element corresponding to the handwriting device list, and forming a verification set of the handwriting device list from the set elements; Establish a screening model, the expression of the screening model is: ； Where: (s, U) is the input data, s is the device information to be written, and U is the verification set; s∈U means that the device information to be written belongs to the verification set; the output data is the verification screening value SU(s, U), and the verification screening value is 1 or 0; Analyze the verification set and the device information to be selected by using the screening model to obtain the verification screening value of the device information to be selected; When the check screening value is 1, the candidate writing device information is eliminated; When the verification screening value is 0, the handwriting device information to be selected is marked as handwriting device information.

4. According to claim 1, a method for intelligent supervision of handwritten answers to subjective questions in a computer-based academic examination, characterized in that: Methods for recommending handwriting device combinations to administrators based on computer-based test information include: A preset handwriting feature table is used to count corresponding handwriting features; Identify computer-based test information, perform feature analysis on the computer-based test information according to the handwriting feature table, obtain handwriting features corresponding to the computer-based test information, and integrate the handwriting features into computer-based test equipment requirements for the computer-based test information; Obtain a list of handwriting devices, analyze the list of handwriting devices according to the requirements of the computer-based test equipment, and obtain a combination of devices to be selected; Screen the candidate device combinations to determine the recommended handwriting device combinations.

5. According to claim 4, a method for intelligent supervision of handwritten answers to subjective questions in computer-based exams, characterized in that: The method for screening candidate device combinations includes: Obtain the computer-based examination conditions and estimate the additional costs of the candidate device combinations according to the computer-based examination conditions. Obtain a handwriting feature table, set the handwriting values of the corresponding handwriting devices for the handwriting features in the handwriting feature table; establish a handwriting value statistical table based on the handwriting features, handwriting devices, and handwriting values. Identify each handwriting device corresponding to the candidate device combination and label it as a unit device; determine the handwriting features that the unit device in the candidate device combination needs to solve according to the computer-based examination device requirements, and match the handwriting values of the corresponding handwriting features from the handwriting value statistical table according to the unit device and the handwriting features. Mark the handwritten value as SY i , i represents the corresponding handwriting feature in the computer-based test equipment requirements, i=1, 2, ..., n, n is the number of handwriting features in the computer-based test equipment requirements; Calculate the screening values of the corresponding candidate device combinations according to the screening formula, and determine the recommended handwriting device combinations according to the screening values. The screening formula is: ； In the formula: SA is the screening value; b1 and b2 are both proportionality coefficients, and the value ranges are 0 < b1 ≤ 1 and 0 < b2 ≤ 1; XB is the additional cost.

6. According to claim 1, a method for intelligent supervision of handwritten answers to subjective questions in a computer-based academic examination, characterized in that: Monitor the anomalies of the answer data to obtain the monitoring results; supplement the monitoring results to the answer data.

7. The intelligent supervision method for handwritten answers to subjective questions in computer-based exams according to claim 6 is characterized in that: The method for monitoring the anomalies of the answer data includes: Set the anomaly monitoring items and the corresponding anomaly criteria; establish an anomaly recognition model according to the anomaly monitoring items and the anomaly criteria. Collect data on the answer data according to the anomaly monitoring items to obtain the monitoring item data corresponding to the anomaly monitoring items. Analyze the monitoring item data through the anomaly recognition model to obtain the corresponding monitoring results.

Images