Answering state evaluation method and related device, learning machine and storage medium

Abstract

The invention discloses an answer state evaluation method, a related device, a learning machine and a storage medium, and the method comprises the steps: obtaining the actual answer duration and answer information of a target object answering a target question, and obtaining a target duration range of the target object theoretically answering the target question; detecting a relationship between the actual answering duration and a target duration range; and in response to the fact that the actual answering duration is not in the target duration range, obtaining an answering state when the target object answers the target question based on the answering information. According to the scheme, a more targeted answer condition can be detected.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a method for evaluating a question answering status and related devices, a learning machine and a storage medium. Background Art

[0002] With the rapid development of smart education-related technologies, software / hardware products used for answering questions and practicing have attracted more and more attention from the market.

[0003] At present, existing software / hardware products usually display the standard answer to the question after the target object answers it, so as to understand the target object's answer situation based on the standard answer. However, the answer situation learned in this way mainly reflects whether the answer is right or wrong, which is relatively general.

[0004] Therefore, how to detect more targeted answers has become an urgent problem to be solved. Summary of the invention

[0005] The main technical problem solved by this application is to provide a method for evaluating the answering status and related devices, equipment and storage media, which can detect more targeted answering situations.

[0006] In order to solve the above technical problems, the first aspect of the present application provides a method for evaluating answering status, which includes: obtaining the actual answering time and answering information of the target object in answering the target question, and obtaining the target time range for the target object to theoretically answer the target question; detecting the relationship between the actual answering time and the target time range; in response to the actual answering time not being within the target time range, obtaining the answering status of the target object when answering the target question based on the answer information.

[0007] In order to solve the above-mentioned technical problems, the second aspect of the present application provides a question answering status evaluation device, which includes an acquisition module, a detection module and an evaluation module; the acquisition module is used to obtain the actual answering time and answer information of the target object in answering the target question, and obtain the target time range for the target object to theoretically answer the target question; the detection module is used to detect the relationship between the actual answering time and the target time range; the evaluation module is used to obtain the answering status of the target object when answering the target question based on the answer information in response to the actual answering time not being within the target time range.

[0008] In order to solve the above technical problems, the third aspect of the present application provides a learning machine, including a memory and a processor coupled to each other, and the processor is used to execute program instructions stored in the memory to implement the answer status evaluation method in the above first aspect.

[0009] In order to solve the above technical problems, the fourth aspect of the present application provides a computer-readable storage medium on which program instructions are stored. When the program instructions are executed by a processor, the method for evaluating the answering status in the above first aspect is implemented.

[0010] The above scheme obtains the actual answering time and answering information of the target object in answering the target question, and obtains the target time range of the target object in theoretical answering the target question, and then detects the relationship between the actual answering time and the target time range, and in response to the actual answering time not being within the target time range, based on the answering information, obtains the answering status of the target object when answering the target question. Therefore, on the one hand, by detecting the relationship between the actual answering time and the target time range, the answering status of the target object when answering the target question can be determined according to the answering information in response to the situation that the actual answering time is not within the target time range, and then the two levels of the actual answering time and the answering information can be organically combined to refine the evaluation dimension of the answering status, and on the other hand, by detecting the answering status of the target object when answering the target question, the answering status is more targeted compared to providing the standard answer to the target question. Therefore, a more targeted answering situation can be detected, that is, the answering status of the target object when answering the target question, and the accuracy of the answering status evaluation can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 This is a flow chart of an embodiment of a method for evaluating the status of answering questions in this application;

[0012] Figure 2 This is a schematic diagram of the framework of an embodiment of a device for evaluating the status of answering questions in this application;

[0013] Figure 3 This is a schematic diagram of the framework of an embodiment of a learning machine of the present application;

[0014] Figure 4 A schematic diagram of a framework of an embodiment of a computer-readable storage medium of the present application. DETAILED DESCRIPTION

[0015] The scheme of the embodiment of the present application is described in detail below in conjunction with the drawings of the specification.

[0016] In the following description, for the purpose of explanation rather than limitation, specific details such as specific system structures, interfaces, and technologies are provided to facilitate a thorough understanding of the present application.

[0017] The term "multiple" herein means two or more than two. In addition, the term "at least one" herein means any combination of at least two of any one or more of a plurality of, for example, including at least one of A, B, and C, and may mean including any one or more elements selected from the set consisting of A, B, and C.

[0018] See also Figure 1 , Figure 1 This is a flow chart of an embodiment of the method for evaluating the answering status of the present application, which specifically includes the following steps:

[0019] Step S11: obtaining the actual answering time and answering information of the target subject in answering the target question, and obtaining the target time range for the target subject to theoretically answer the target question.

[0020] In the disclosed embodiment, the target objects may include students, practitioners in professional fields, and other objects who need to take exams.

[0021] In the disclosed embodiment, the target questions may include, but are not limited to, subjective questions, objective questions, etc. In addition, the target questions may include subject areas such as mathematics, foreign languages, Chinese, physics, etc., and may also include professional skill testing areas such as computers, economics, accounting, and law.

[0022] In one implementation scenario, the target subject can directly answer the target question on the screen of the learning machine or the device running the question-answering application. By obtaining the target subject's answer record on the learning machine or the device running the question-answering application, the actual answering time and answer information of the target subject can be obtained. In this case, as a possible example, the answer information may include but is not limited to: behavior record information, answer writing track on the answer sheet, answer word count, etc.

[0023] In another implementation scenario, the target subject can also answer the target question on paper, take a photo of the answer paper and upload it to a learning machine or a device running a question-solving application, so as to detect and identify the captured image, obtain the target subject's answer information for the target question, and use the built-in timer of the learning machine or the device running the question-solving application to count the time from the target subject starting the timing to receiving the uploaded photo as the actual answering time of the target subject, or a separate timing device such as a watch can be used to obtain the actual answering time of the target subject. In this case, as a possible example, the answer information may include but is not limited to: the writing track of the answer on the answer sheet, the number of words in the answer, and other information.

[0024] In an implementation scenario, the target time range can be obtained by correcting the preset time range of the target question by the historical answer time. Specifically, the preset time range and the historical answer time of the target question can be obtained, and the historical answer time of the reference object at the same learning level as the target object can be selected from the historical answer time, and the preset time range can be corrected to obtain the target time range. Among them, the preset time range represents the time range theoretically required to answer the target question, such as can be manually or machine-determined based on the difficulty, question type, knowledge field and other relevant information of the target question; in addition, the historical answer time represents the actual answer time of the reference object who has answered the target question for the target question, which can be obtained by online collection or acquisition from a database; in addition, the learning level is used to reflect the learning ability, such as can include excellent, medium, poor or other more refined classification levels. Therefore, based on the historical answering time of reference objects at the same learning level as the target object, the preset time range theoretically required to answer the target question is corrected to obtain the target time range for the target object to theoretically answer the target question. This takes into account the target object's learning ability and formulates the target time range in a targeted manner, which helps to improve the flexibility of the target time range.

[0025] In a specific implementation scenario, the preset duration range can be a bounded interval including a minimum duration and a maximum duration, such as 40 seconds to 50 seconds, or it can be a unilateral unbounded interval including a minimum duration or a maximum duration, such as not less than 40 seconds or not more than 50 seconds. Accordingly, the target duration range obtained by correcting the preset duration range through the historical answer duration of a reference object at the same learning level as the target object can also be a bounded interval or a unilateral unbounded interval.

[0026] Exemplarily, taking the preset duration range as a bounded interval as an example, the minimum duration is 40 seconds and the maximum duration is 50 seconds, that is, the preset duration range is 40 seconds to 50 seconds. If the learning level of the target object is excellent, a representative duration such as 35 seconds is obtained through the historical answering duration of the reference object in the excellent level. The preset duration range can be corrected to obtain a target duration range as a bounded interval such as 30 seconds to 40 seconds, or, the target duration range is corrected to be a unilateral unbounded interval such as less than or equal to 35 seconds. It should be noted that the preset duration range will vary with the target topic, and the historical answering duration will vary with the target topic or the learning level of the target object. Therefore, the above-mentioned specific values ​​about the duration are only used as examples and do not constitute a specific limitation on the present application. In addition, the representative duration can be the average value of the historical answering duration of multiple reference objects at the same level, or it can be the duration with the largest number of the same durations in the historical answering durations of multiple reference objects at the same level. The method for obtaining the representative duration is not specifically limited here.

[0027] In a specific implementation scenario, the actual answering time of the current target object can be stored in the database as the historical answering time of other target objects when answering the target questions, so as to increase the data source of historical answering time for the database in real time.

[0028] In a specific implementation scenario, before obtaining the historical answering time, the actual answering time of the reference subjects who have answered the target question can be screened, and then the actual answering time of the screened reference subjects can be used as the historical answering time, thereby improving the validity of the historical answering time as reference data. For example, if the actual answering time of a reference subject who has answered the target question is too short to be used as a valid time, the actual answering time of the reference subject can be eliminated.

[0029] Step S12: Detect the relationship between the actual answering time and the target time range.

[0030] In one implementation scenario, the target duration range is a bounded interval, and the target duration range includes a minimum duration and a maximum duration. When the actual answering duration is less than the minimum duration or greater than the maximum duration, it indicates that the actual answering duration is not within the target duration range. When the actual answering duration is greater than or equal to the minimum duration and less than or equal to the maximum duration, it indicates that the actual answering duration is within the target duration range. For the specific form of the bounded interval, please refer to the relevant examples in the above embodiments, which will not be repeated here.

[0031] In another implementation scenario, the target duration range is a unilateral unbounded interval, and the target duration range may be not less than the minimum duration or not greater than the maximum duration. When the actual answering duration is less than the minimum duration or greater than the maximum duration, it indicates that the actual answering duration is not within the target duration range. When the actual answering duration is greater than or equal to the minimum duration or less than or equal to the maximum duration, it indicates that the actual answering duration is within the target duration range. For the specific form of the unilateral unbounded interval, please refer to the relevant examples in the above embodiments, which will not be repeated here.

[0032] Step S13: In response to the actual answering time not being within the target time range, the answering status of the target subject when answering the target question is obtained based on the answering information.

[0033] In the disclosed embodiments, when the actual answering time is not within the target time range, the answering status of the target object when answering the target question is obtained based on the answering information. For the case where the actual answering time is not within the target time range, the two levels of actual answering time and answering information can be organically combined to refine the evaluation dimension of the answering status and improve the efficiency and accuracy of the answering status evaluation. In addition, when the actual answering time is within the target time range, the answering status of the target object can also be obtained through the answering information, and the evaluation process of the answering status of the target object can be similar to the specific process of obtaining the answering status of the target object when answering the target question based on the answering information when the actual answering time is not within the target time range in the following embodiments. That is to say, regardless of whether the actual answering time is within the target time range or not, the relationship between the actual answering time and the target time range can be used as one of the evaluation dimensions of the answering status, and then combined with the evaluation dimensions of the answering status obtained through the subsequent answering information to obtain the final answering status. For example, the score representing the final answering status can be obtained by calculating the scores of each evaluation dimension, or the answering status of the most representative evaluation dimension can be selected as the final answering status based on the specific test results. The specific combination of each evaluation dimension is not limited here. Of course, when the actual answering time is within the target time range, it can also be directly assumed that the target object is in a normal answering state when answering the target question. The above examples are only a few possible examples in actual application. The processing method for the actual answering time within the target time range is not limited here.

[0034] In one implementation scenario, the answer information may include the answer writing track on the answer sheet. In order to obtain the answer status of the target object when answering the target question, the answer writing track can be recognized by text recognition to determine whether the answer writing track contains the target writing track that is not related to the answer, and based on whether the answer writing track contains the target writing track, the answer status of the target object when answering the target question is determined. For example, when the answer writing track contains the target writing track, the evaluation dimension of the answer writing track can determine that the answer status of the target object when answering the target question is an abnormal answer. Therefore, by performing text recognition on the answer writing track and taking whether the answer writing track contains the target writing track that is not related to the answer as a dimension for answer status evaluation, the answer status can be determined according to the situation of the target object answering the target question on the paper, which helps to further refine the evaluation dimension of the answer status and improve the precision of the answer status evaluation.

[0035] In a specific implementation scenario, the recognition result of text recognition of the answer writing trajectory includes at least one of the answer text and the answer line, wherein the answer line may include a straight line segment, a curve or a shape composed of lines, etc., and the form of the answer line is not specifically limited here. Correspondingly, the target writing trajectory includes at least one of the target text and the target line, wherein the target text is the target writing trajectory when the recognition result includes the answer text, and the target line is the target writing trajectory when the recognition result includes the answer line, and then when the recognition result only includes the answer text or only includes the answer line or includes both the answer text and the answer line, determine whether the answer writing trajectory contains target text and / or target lines that are not related to the answer. Therefore, by adopting different strategies for determining the target writing trajectory based on the different contents included in the recognition result of text recognition of the answer writing trajectory, the accuracy and flexibility of evaluating the answering status of the target object can be improved.

[0036] In a specific implementation scenario, when the recognition result of the text recognition of the answer writing track contains the answer text, in order to determine whether the answer writing track contains the target writing track that is not related to the answer, the answer text in the recognition result can be analyzed to determine whether the answer text in the recognition result contains the target text that is not related to the answer, wherein the target text is the target writing track when the answer text is included. Exemplarily, the target question is a reading comprehension question of the Chinese subject, and the answer text in the recognition result is analyzed to determine that the answer text contains mathematical formula text. Since the mathematical formula text is not related to the answer to the current reading comprehension question, it is determined that the answer text in the recognition result contains the target text that is not related to the answer, and then the answer status of the target object can be determined as an abnormal answer.

[0037] In a specific implementation scenario, when the recognition result of the text recognition of the answer writing track contains the answer line, in order to determine whether the answer writing track contains the target writing track that is not related to the answer, at least the target question and the answer line can be analyzed to determine whether the answer line in the recognition result contains the target line that is not related to the answer, wherein the target line is the target writing track when the answer line is included. It should be noted that the answer line included in the answer writing track may completely belong to the target line that is not related to the answer. For example, when the target question is a Chinese reading comprehension question or a foreign language translation question that only involves text, if the answer writing track includes the answer line, the answer line can be directly determined as the target line that is not related to the answer; the answer line included in the answer writing track may also partially belong to the target line that is not related to the answer, or may not belong to the target line at all. For example, when the target question is a mathematical geometry question or a physical mechanics analysis question that may involve answer lines related to the answer, it is necessary to specifically judge whether the answer line in the recognition result contains the target line that is not related to the answer. Exemplarily, the target question is a math geometry test question, the target question includes question text and geometric figures, the answer writing track includes answer lines and answer text, by analyzing the target question and the answer line, when the answer line is an auxiliary line located on the geometric figure, it can be determined that the answer line is an answer line related to the answer. Specifically, it can be determined whether the answer line is an auxiliary line in combination with the reference answer to the target question. When the answer line does not intersect with the geometric figure or the answer text does not contain relevant instructions about the answer line, it can be determined that the answer line is a target line unrelated to the answer, and then the answer status of the target object can be determined as an abnormal answer.

[0038] In a specific implementation scenario, when the recognition result of the text recognition of the answer writing track contains the answer text and the answer line, in order to determine whether the answer writing track contains the target writing track that is not related to the answer, the answer text in the recognition result can be analyzed to determine whether the answer text in the recognition result contains the target text that is not related to the answer, and at least the target question and the answer line can be analyzed to determine whether the answer line in the recognition result contains the target line that is not related to the answer. When the answer text contains the target text and / or the answer line includes the target line, it can be determined that the answer status of the target object is an abnormal answer. Among them, the target text is the target writing track when it contains the answer text, and the target line is the target writing track when it contains the answer line.

[0039] In a specific implementation scenario, the recognition result of the text recognition of the answer writing track may also include the case of being unable to be recognized. For example, when the text recognition of the answer writing track cannot be classified as the answer text or the answer line, the recognition result is determined to be unrecognizable. And in the case of the recognition result being unrecognizable, it can be directly obtained that the answer status of the target subject when answering the target question is abnormal.

[0040] In one implementation scenario, the answer information may include the number of words in the answer sheet. In order to obtain the answer status of the target subject when answering the target question, the word count of the answer sheet can be performed. When the word count obtained is less than the preset word count, the answer status of the target subject when answering the target question is determined to be an abnormal answer. Among them, the preset word count can be determined by the word count of the reference answer of the target question, or it can be determined according to the question type of the target question. The method for determining the preset word count is not limited here. Therefore, using the number of words in the answer sheet as a dimension for evaluating the answer status can further improve the precision of the answer status evaluation.

[0041] In an implementation scenario, before obtaining the answering status of the target object when answering the target question based on the answer information, the learning level of the target object can also be obtained. The learning level is used to reflect the learning ability, such as can include excellent, medium, poor or other more refined level classifications, and then the answering status of the target object when answering the target question can be obtained according to the learning level and answer information of the target object. As a possible example, when it is determined that the learning level of the target object is medium or poor, the answering status of the target object when answering the target question is obtained through the answer writing track in the above embodiment, and / or, the answering status of the target object when answering the target question is obtained through the behavior record information. As another possible example, when it is determined that the learning level of the target object is excellent, the answering status of the target object when answering the target question can be obtained only through the behavior record information in the following embodiment. Among them, the specific process of obtaining the answering status of the target object when answering the target question through the behavior record information can be referred to the following embodiment, which will not be repeated here. Therefore, based on the target object's learning level and answer information, the answer status of the target object when answering the target question is obtained, which can improve the accuracy of the answer status evaluation. Moreover, when the target object's learning level is excellent, the step of using the answer writing trajectory as the answer status evaluation standard can be omitted, which can improve the efficiency of the answer status evaluation to a certain extent.

[0042] In one implementation scenario, the answer status evaluation method is executed by a learning machine, and the answer information may include behavior record information, and the behavior record information is collected by the learning machine from the target object, and then the answer status of the target object can be obtained through the behavior record information. Therefore, using the behavior record information collected by the learning machine from the target object as a dimension of the answer status evaluation can refine the evaluation dimension of the answer status from the behavioral level of the target object's process of answering the target question.

[0043] In a specific implementation scenario, the answer information may include the answer writing track on the answer sheet, and the behavior record information includes the location record information of the track erasing behavior. Then, the location record information can be used to determine whether the target object has repeated erasing actions during the answering process, and in the case of repeated erasing actions, the target object's answering status can be obtained through the relative position relationship between the occurrence position of the repeated erasing action and the answer writing track. For example, in the case where the occurrence position of the repeated erasing action does not intersect with the position of the answer writing track, the target object's answering status can be obtained as abnormal answering in the evaluation dimension of the behavior record information. Among them, the track erasing behavior can be directly realized by manipulating the relevant functions on the stylus matched with the learning machine. In addition, in order to further improve the accuracy of the judgment of repeated erasing actions, it is also possible to combine the situation of whether the target object selects the erasing option on the answering interface of the learning machine with the location record information of the track erasing behavior to determine whether the target object has repeated erasing actions during the answering process. For example, if the target object selects the erasing option and repeatedly moves in a certain position interval, it can be determined that the target object has repeated erasing actions during the answering process. Therefore, through the relative position relationship between the occurrence location of the repeated erasing action and the answer writing trajectory, it is possible to determine whether the repeated erasing action is erasing in a normal answering state or erasing in an abnormal answering state, so as to classify the behavior record information more finely, thereby further improving the granularity of the answering status evaluation.

[0044] In a specific implementation scenario, the answer information may include the answer writing track on the answer sheet, and the answer writing track is formed by the target object manipulating the stylus to write on the learning machine. The behavior record information includes the manipulation record information of the stylus, and then the answer status of the target object can be obtained by detecting whether the answer writing track matches the manipulation record information. For example, in the case where a large section of the answer writing track is obtained by copying and pasting operations, it is determined that the answer writing track does not match the manipulation record information. For another example, when the target object is in the process of answering the target question, there is a manipulation record information of switching the background, then it is determined that the answer writing track does not match the manipulation record information, and then when the answer writing track does not match the manipulation record information, the target object's answer status is obtained as an abnormal answer. Therefore, by judging whether the manipulation record information of the entire process of the target object answering the target question matches the final answer writing track, it can be determined whether there is an abnormal operation record in the process of the target object answering the target question, and whether there is an abnormal operation record is used as a dimension for the evaluation of the answer status, which helps to improve the granularity of the answer status evaluation.

[0045] In a specific implementation scenario, the behavior record information may include video record information taken of the target object when the target object is answering and audio record information collected from the answering environment, and then the first answering state detected by the target object in the video mode can be obtained through the video record information, and the audio record information can be detected to obtain the second answering state detected by the target object in the audio mode, and then the answering state of the target object can be obtained through the first answering state and the second answering state. It should be noted that before formally collecting the video mode and audio mode of the target object, this application has clearly informed the personal information processing rules and obtained the individual's voluntary consent. For example, in the case of using obvious signs / information to inform the personal information processing rules, personal authorization is obtained through pop-up information or by asking the individual to upload his or her personal information. Therefore, combining the video record information taken of the target object when the target object is answering and the audio record information collected from the answering environment to assist in analyzing the answering state of the target object can improve the accuracy of the answering state evaluation.

[0046] As a possible implementation method, the first answering state and the second answering state can be scores representing whether the answering state is abnormal, and the first answering state and the second answering state can be fused by the scores to obtain the score representing the final answering state. In addition, different weight values ​​can be set for the first answering state and the second answering state, and then when the score representing the final answering state is obtained by score fusion, the weight of the first answering state can be set to be greater or less than the weight of the second answering state according to different situations, so as to improve the flexibility of answering state evaluation.

[0047] As another possible implementation, the first answering state and the second answering state can also be more specific detection results. For example, the detection result of the first answering state can be whether the target object has behaviors such as inattention or long-term eye closure in the video mode, and the detection result of the second answering state can be whether the target object chats or plays other background audio unrelated to the answer in the audio mode, and then the final answering state can be obtained by comprehensively judging the specific detection results of the first answering state and the second answering state. As a possible example, when the detection result of the audio mode includes background audio, and the detection result of the video mode does not contain abnormal answers, the final answering state can be considered to be a normal answer.

[0048] In an implementation scenario, in addition to obtaining the answer status of the target object when answering the target question based on the answer information, the answer information can also be corrected to obtain the correction result of the target object answering the target question. Exemplarily, a specific correction model can be pre-trained through sample questions, sample answers and sample answer information of the sample object, and the sample questions are marked with the real correction results of the sample answer information. The correction model is used to analyze the sample answer information of the sample object to obtain the predicted correction result, so that the network parameters of the correction model can be adjusted based on the difference between the predicted correction result and the real correction result, so that the correction result output by the correction model can be made as accurate as possible. Of course, the above example is only a possible example in the actual application process, and does not limit the specific method of obtaining the correction result. Therefore, while obtaining the answer status of the target object when answering the target question based on the answer information, the correction result of the target object answering the target question is obtained, which can enrich the evaluation dimension of the target object answering the target question, so as to reflect the learning or test status of the target object as realistically and comprehensively as possible.

[0049] In an implementation scenario, after obtaining the answering status of the target object when answering the target question, a prompt message can be generated through the answering status, answering information and the correction result of the target question, and the prompt message can be sent to the associated object of the target object. For example, if the target object is a student, the associated object of the student can be a parent; for another example, if the target object is an interviewee, the associated object can be the interviewee himself. In addition, the form of the prompt message can include but is not limited to the form of pop-up window reminder viewing or direct display on the interface. Therefore, the prompt message generated according to the answering status, answering information and the correction result of the target question is sent to the associated object of the target object, which can remind the associated object in time to assist in adjusting the learning status or test status of the target object.

[0050] The above scheme obtains the actual answering time and answering information of the target object in answering the target question, and obtains the target time range of the target object in theoretical answering the target question, and then detects the relationship between the actual answering time and the target time range, and in response to the actual answering time not being within the target time range, based on the answering information, obtains the answering status of the target object when answering the target question. Therefore, on the one hand, by detecting the relationship between the actual answering time and the target time range, the answering status of the target object when answering the target question can be determined according to the answering information in response to the situation that the actual answering time is not within the target time range, and then the two levels of the actual answering time and the answering information can be organically combined to refine the evaluation dimension of the answering status, and on the other hand, by detecting the answering status of the target object when answering the target question, the answering status is more targeted compared to providing the standard answer to the target question. Therefore, a more targeted answering situation can be detected, that is, the answering status of the target object when answering the target question, and the accuracy of the answering status evaluation can be improved.

[0051] See also Figure 2 , Figure 2 This is a framework diagram of an embodiment of the answering status evaluation device of the present application, the answering status evaluation device 20 includes an acquisition module 21, a detection module 22 and an evaluation module 23, the acquisition module 21 is used to obtain the actual answering time and answer information of the target object in answering the target question, and obtain the target time range for the target object to theoretically answer the target question; the detection module 22 is used to detect the relationship between the actual answering time and the target time range; the evaluation module 23 is used to obtain the answering status of the target object when answering the target question based on the answer information in response to the actual answering time not being within the target time range.

[0052] In the above scheme, the answering status evaluation device 20 obtains the actual answering time and answering information of the target object in answering the target question, and obtains the target time range of the target object in theoretical answering the target question, and then detects the relationship between the actual answering time and the target time range, and in response to the actual answering time not being within the target time range, based on the answering information, obtains the answering status of the target object when answering the target question. Therefore, on the one hand, by detecting the relationship between the actual answering time and the target time range, the answering status of the target object when answering the target question can be determined according to the answering information in response to the situation that the actual answering time is not within the target time range, and then the two levels of the actual answering time and the answering information can be organically combined to refine the evaluation dimension of the answering status, and on the other hand, by detecting the answering status of the target object when answering the target question, the answering status is more targeted compared to providing the standard answer of the target question. Therefore, a more targeted answering situation can be detected, that is, the answering status of the target object when answering the target question, and the accuracy of the answering status evaluation can be improved.

[0053] In some disclosed embodiments, the evaluation module 23 includes a first evaluation sub-module, which is used to perform text recognition based on the answer writing trajectory to determine whether the answer writing trajectory contains a target writing trajectory that is not related to the answer; based on whether the answer writing trajectory contains the target writing trajectory, determine the answer status of the target object when answering the target question.

[0054] Therefore, by performing text recognition on the answer writing trajectory and taking whether the answer writing trajectory contains target writing trajectories that are not related to the answer as a dimension for evaluating the answer status, the answer status can be determined based on the paper situation of the target subject answering the target question, which helps to further refine the evaluation dimension of the answer status and improve the precision of the answer status evaluation.

[0055] In some disclosed embodiments, the first evaluation submodule includes a recognition unit and a first response unit and a second response unit, the recognition unit being used to perform text recognition based on the answer writing trajectory to obtain a recognition result; wherein the recognition result includes at least one of answer text and answer lines; the first response unit is used to respond to the recognition result containing answer text, analyze based on the answer text in the recognition result, and determine whether the answer text in the recognition result contains target text that is not related to the answer; wherein the target text is the target writing trajectory when the answer text is included; the second response unit is used to respond to the recognition result containing answer lines, analyze at least based on the target question and the answer lines, and determine whether the answer lines in the recognition result contain target lines that are not related to the answer; wherein the target lines are the target writing trajectory when the answer lines are included.

[0056] Therefore, by taking into account the different contents included in the recognition results of the text recognition of the answer writing trajectory and adopting different strategies for determining the target writing trajectory, the accuracy and flexibility of evaluating the answering status of the target object can be improved.

[0057] In some disclosed embodiments, the evaluation module 23 includes a second evaluation sub-module, the answer status evaluation method is executed by a learning machine, the answer information includes behavior record information, and the behavior record information is collected by the learning machine for the target object, and the second evaluation sub-module is used to obtain the answer status of the target object based on the behavior record information.

[0058] Therefore, using the behavior record information collected by the learning machine on the target object as a dimension of the answering evaluation state can refine the evaluation dimension of the answering state from the behavioral level of the target object's process of answering the target question.

[0059] In some public embodiments, the answer information also includes the answer writing track on the answer sheet, and the behavior record information includes the position record information of the track erasing behavior. The second evaluation submodule is specifically used to determine whether the target object has repeated erasing actions in the answering process based on the position record information; in response to the existence of repeated erasing actions, the answering status of the target object is obtained based on the relative position relationship between the occurrence position of the repeated erasing action and the answer writing track.

[0060] Therefore, through the relative position relationship between the occurrence location of the repeated erasing action and the answer writing trajectory, it is possible to determine whether the repeated erasing action is erasing in a normal answering state or erasing in an abnormal answering state, so as to classify the behavior record information more finely, thereby further improving the granularity of the answering status evaluation.

[0061] In some public embodiments, the answer information also includes an answer writing track on the answer sheet, the answer writing track is formed by the target object controlling the stylus to write on the learning machine, the behavior record information includes the control record information of the stylus, and the second evaluation submodule is specifically used to detect whether the answer writing track matches the control record information; in response to whether the answer writing track matches the control record information, the answer status of the target object is obtained.

[0062] Therefore, by judging whether the manipulation record information of the entire process of the target object answering the target question matches the final answer writing trajectory, it is possible to determine whether there are abnormal operation records in the process of the target object answering the target question, and using the existence of abnormal operation records as a dimension for answering status evaluation, which helps to improve the granularity of answering status evaluation.

[0063] In some public embodiments, the behavior recording information includes video recording information of the target object when the target object is answering the questions and audio recording information collected from the answering environment. The second evaluation submodule is specifically used to obtain a first answering state detected by the target object in the video mode based on the video recording information, and to perform detection based on the audio recording information to obtain a second answering state detected by the target object in the audio mode; based on the first answering state and the second answering state, the answering state of the target object is obtained.

[0064] Therefore, combining the video recording information of the target object when answering the questions and the audio recording information collected from the answering environment to assist in analyzing the answering status of the target object can improve the accuracy of the answering status evaluation.

[0065] In some disclosed embodiments, the acquisition module 21 is specifically used to obtain a preset time range and a historical answer time of a target question; wherein the historical answer time represents the actual answer time of a reference object who has answered the target question, and the preset time range represents the time range theoretically required to answer the target question; based on the historical answer time of a reference object at the same learning level as the target object, the preset time range is corrected to obtain a target time range; wherein the learning level is used to reflect learning ability.

[0066] Therefore, based on the historical answering time of reference objects at the same learning level as the target object, the preset time range theoretically required to answer the target question is corrected to obtain the target time range for the target object to theoretically answer the target question. This takes into account the target object's learning ability and formulates the target time range in a targeted manner, which helps to improve the flexibility of the target time range.

[0067] In some disclosed embodiments, the answering status evaluation device 20 also includes a generation module and a sending module. The generation module is used to generate a prompt message based on the answering status, answer information and correction results of the target question; the sending module is used to send the prompt message to the associated objects of the target object.

[0068] Therefore, the prompt message generated according to the answering status, answering information and correction result of the target question is sent to the associated object of the target object, which can remind the associated object in time to assist in adjusting the learning status or examination status of the target object.

[0069] See also Figure 3 , Figure 3 The learning machine 30 includes a memory 31 and a processor 32 coupled to each other, and the processor 32 is used to execute program instructions stored in the memory 31 to implement the steps in any of the above-mentioned answer status evaluation method embodiments.

[0070] Specifically, the processor 32 is used to control itself and the memory 31 to implement the steps of any of the above-mentioned answer status evaluation method embodiments. The processor 32 can also be called a CPU (Central Processing Unit). The processor 32 may be an integrated circuit chip with signal processing capabilities. The processor 32 can also be a general-purpose processor, a digital signal processor (Digital Signal Processor, DSP), an application-specific integrated circuit (Application Specific Integrated Circuit, ASIC), a field-programmable gate array (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. In addition, the processor 32 can be implemented by an integrated circuit chip.

[0071] In one implementation scenario, the learning machine may include a video acquisition device and an audio acquisition device. The video acquisition device is used to collect video recording information when the target object answers the target question, and the audio acquisition device is used to collect audio recording information when the target object answers the target question. Before formally collecting the video modality and audio modality of the target object, the personal information processing rules are clearly informed and the individual's voluntary consent is obtained. For example, the personal information processing rules are notified by obvious signs / information, and the individual's authorization is obtained through pop-up information or by asking the individual to upload his or her personal information.

[0072] In the above scheme, the learning machine 30 obtains the actual answering time and answering information of the target object in answering the target question, and obtains the target time range of the target object in theoretical answering the target question, and then detects the relationship between the actual answering time and the target time range, and in response to the actual answering time not being within the target time range, obtains the answering status of the target object when answering the target question based on the answering information. Therefore, on the one hand, by detecting the relationship between the actual answering time and the target time range, the answering status of the target object when answering the target question can be determined according to the answering information in response to the situation that the actual answering time is not within the target time range, and then the two levels of the actual answering time and the answering information can be organically combined to refine the evaluation dimension of the answering status, and on the other hand, by detecting the answering status of the target object when answering the target question, the answering status is more targeted compared to providing the standard answer to the target question. Therefore, a more targeted answering situation can be detected, that is, the answering status of the target object when answering the target question, and the accuracy of the answering status evaluation can be improved.

[0073] See also Figure 4 , Figure 4The schematic diagram of the framework of an embodiment of a computer-readable storage medium 40 of the present application. The computer-readable storage medium 40 stores program instructions 41 that can be executed by a processor, and the program instructions 41 are used to implement the steps of the embodiment of the above-mentioned method for evaluating the answering status.

[0074] In the above scheme, the computer-readable storage medium 40 obtains the actual answering time and answering information of the target object in answering the target question, and obtains the target time range of the target object's theoretical answering of the target question, and then detects the relationship between the actual answering time and the target time range, and in response to the actual answering time not being within the target time range, based on the answering information, obtains the answering state of the target object when answering the target question. Therefore, on the one hand, by detecting the relationship between the actual answering time and the target time range, the answering state of the target object when answering the target question can be determined according to the answering information in response to the situation that the actual answering time is not within the target time range, and then the two levels of the actual answering time and the answering information can be organically combined to refine the evaluation dimension of the answering state, and on the other hand, by detecting the answering state of the target object when answering the target question, the answering state is more targeted compared to providing the standard answer of the target question. Therefore, a more targeted answering situation can be detected, that is, the answering state of the target object when answering the target question, and the accuracy of the answering state evaluation can be improved.

[0075] In some embodiments, the functions or modules included in the device provided by the embodiments of the present disclosure can be used to execute the method described in the above method embodiments. The specific implementation can refer to the description of the above method embodiments, and for the sake of brevity, it will not be repeated here.

[0076] The above description of various embodiments tends to emphasize the differences between the various embodiments. The same or similar aspects can be referenced to each other, and for the sake of brevity, they will not be repeated herein.

[0077] In the several embodiments provided in the present application, it should be understood that the disclosed methods and devices can be implemented in other ways. For example, the device implementation described above is only schematic. For example, the division of modules or units is only a logical function division. There may be other division methods in actual implementation, such as units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, and the indirect coupling or communication connection of devices or units can be electrical, mechanical or other forms.

[0078] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0079] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) or a processor (processor) to perform all or part of the steps of each implementation method of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program code.

[0080] If the technical solution of this application involves personal information, the product using the technical solution of this application has clearly informed the personal information processing rules and obtained the individual's voluntary consent before processing the personal information. If the technical solution of this application involves sensitive personal information, the product using the technical solution of this application has obtained the individual's separate consent before processing the sensitive personal information, and at the same time meets the "explicit consent" requirement. For example, on personal information collection devices such as cameras, clear and prominent signs are set to inform that the personal information collection scope has been entered and personal information will be collected. If the individual voluntarily enters the collection scope, it is deemed that he or she agrees to the collection of his or her personal information; or on the device that processes personal information, the personal information processing rules are notified by obvious signs / information, and the individual's authorization is obtained through pop-up information or by asking the individual to upload his or her personal information; among them, the personal information processing rules may include information such as the personal information processor, the purpose of personal information processing, the processing method, and the type of personal information processed.

Claims

1. A method for evaluating the answering status, characterized in that: include: Obtaining the actual answering time and answering information of the target subject in answering the target question, and obtaining the target time range for the target subject to theoretically answer the target question; Detecting the relationship between the actual answering time and the target time range; In response to the actual answering time not being within the target time range, an answering status of the target subject when answering the target question is obtained based on the answering information.

2. The method according to claim 1, characterized in that The answer information includes the answer writing track on the answer sheet, and obtaining the answer status of the target subject when answering the target question based on the answer information includes: Performing text recognition based on the answer writing track to determine whether the answer writing track contains a target writing track that is not related to the answer; Based on whether the answer writing track contains the target writing track, the answering state of the target object when answering the target question is determined.

3. The method according to claim 2, characterized in that The performing of text recognition based on the answer writing track to determine whether the answer writing track contains a target writing track irrelevant to the answer includes: Performing text recognition based on the answer writing track to obtain a recognition result; wherein the recognition result includes at least one of the answer text and the answer line; In response to the recognition result including the answer text, analyzing the answer text in the recognition result to determine whether the answer text in the recognition result includes target text that is irrelevant to the answer; wherein the target text is the target writing trajectory when the answer text is included; In response to the recognition result containing the answer line, an analysis is performed at least based on the target question and the answer line to determine whether the answer line in the recognition result contains a target line that is not related to the answer; wherein the target line is the target writing trajectory when the answer line is included.

4. The method according to claim 1, characterized in that: The answer status evaluation method is executed by a learning machine, the answer information includes behavior record information, and the behavior record information is collected by the learning machine from the target object, and the answer status of the target object when answering the target question based on the answer information is obtained, including: Based on the behavior record information, the answering status of the target object is obtained.

5. The method according to claim 4, characterized in that The answer information also includes the answer writing track on the answer sheet, the behavior record information includes the location record information of the track erasing behavior, and the answer status of the target object is obtained based on the behavior record information, including: Based on the location record information, determining whether the target object has repeated erasing actions during the question answering process; In response to the existence of the repeated erasing action, the answering state of the target object is obtained based on the relative position relationship between the occurrence position of the repeated erasing action and the answer writing track.

6. The method according to claim 4, characterized in that The answer information also includes an answer writing track on the answer sheet, the answer writing track is formed by the target object controlling the stylus to write on the learning machine, the behavior record information includes the control record information of the stylus, and the answer status of the target object is obtained based on the behavior record information, including: Detecting whether the answer writing track matches the manipulation record information; In response to whether the answer writing track matches the manipulation record information, the answer status of the target object is obtained.

7. The method according to claim 4, characterized in that The behavior record information includes video record information of the target object when the target object is answering and audio record information collected from the answering environment. The answering status of the target object is obtained based on the behavior record information, including: Based on the video recording information, a first answering state detected by the target object in the video mode is obtained, and based on the audio recording information, a second answering state detected by the target object in the audio mode is obtained; Based on the first answering status and the second answering status, the answering status of the target object is obtained.

8. The method according to claim 1, characterized in that The step of obtaining a target time range for the target subject to theoretically answer the target question includes: Obtaining a preset time range and a historical answer time of the target question; wherein the historical answer time represents the actual answer time of the reference subject who has answered the target question to the target question, and the preset time range represents the time range theoretically required to answer the target question; Based on the historical answering time of a reference object at the same learning level as the target object, the preset time range is corrected to obtain the target time range; wherein the learning level is used to reflect the learning ability.

9. The method according to claim 1, characterized in that: After obtaining the answer status of the target subject when answering the target question based on the answer information, the method further includes: Generate a prompt message based on the answering status, the answering information and the correction result of the target question; The prompt message is sent to the associated objects of the target object.

10. A device for evaluating the state of answering questions, characterized in that: include: An acquisition module, used to acquire the actual answering time and answering information of the target subject in answering the target question, and to acquire the target time range for the target subject to theoretically answer the target question; A detection module, used to detect the relationship between the actual answering time and the target time range; An evaluation module is used to obtain, in response to the actual answering time not being within the target time range, an answering status of the target subject when answering the target question based on the answering information.

11. A learning machine, characterized in that: It includes a memory and a processor coupled to each other, and the processor is used to execute program instructions stored in the memory to implement the answer status evaluation method described in any one of claims 1 to 9.

12. A computer-readable storage medium having program instructions stored thereon, characterized in that: When the program instructions are executed by the processor, the method for evaluating the answering status described in any one of claims 1 to 9 is implemented.

Images