Method for judging type of question asked by teacher, and method and device for analyzing question asked by classroom

Abstract

The invention discloses a teacher question type judgment method and device and a classroom question analysis method and device. The method for judging the question type of the teacher comprises the following steps of: obtaining to-be-judged content which is obtained by text data obtained by converting voice of the teacher into a text; the to-be-judged content is matched with a preset rule set, a hit rule is obtained, the preset rule set comprises a plurality of preset rules, each preset rule has a corresponding question type, and each preset rule has a corresponding priority; and taking the rule with the highest priority in the hit rules as the rule corresponding to the to-be-judged content, and extracting the question type corresponding to the rule as the question type of the to-be-judged content. The method is simple and practical, and fits the actual scene of teaching.

Description

Technical Field

[0001] The present invention relates to the technical field of multimedia processing, and more specifically, to a method for judging the type of teacher's questions, a method for analyzing classroom questions, a device, a device and a storage medium. Background Art

[0002] With the development of technology, in the information technology environment, people increasingly use multimedia technology to identify each link in classroom teaching, and then obtain the evaluation results of classroom teaching by analyzing each link. Among them, the evaluation of teacher-student interaction is an important part of classroom teaching evaluation, and teacher's questions are an important link in the teacher-student interaction link.

[0003] However, in the related art, there is a lack of a specific method for evaluating classroom teaching based on teacher's questions. Summary of the Invention

[0004] In order to solve the technical problem of the lack of a specific method for evaluating classroom teaching based on teacher's questions in the prior art, the present invention provides a method. The technical solution adopted by the present invention is as follows.

[0005] In a first aspect, the present invention provides a method for judging the type of teacher's questions, including:

[0006] Obtain the content to be determined, where the content to be determined is obtained from the text data converted from the teacher's voice to text;

[0007] Match the content to be determined with a preset rule set to obtain the rules that are hit. Among them, there are multiple preset rules in the preset rule set, each preset rule has its own corresponding question type, and each preset rule has its own corresponding priority;

[0008] Use the rule with the highest priority among the hit rules as the rule corresponding to the content to be determined, and extract the corresponding question type as the question type of the content to be determined.

[0009] In an implementation manner, the preset rules include: key strings, the logical relationship between key strings, and the order of precedence.

[0010] In a second aspect, the present invention provides a method for analyzing classroom questions, further including:

[0011] Obtain the sentence pattern judgment result of the text data converted from the teacher's voice to text, where the sentence pattern judgment result is obtained through a semantic prediction model based on deep learning;

[0012] Extract the interrogative sentences in the text data according to the sentence pattern judgment result, and determine the extracted interrogative sentences as the content to be determined;

[0013] Using the method for judging the types of teachers' questions described above, obtain the question types of each content to be judged;

[0014] Calculate the proportion of higher-order thinking question types in the question types to obtain the proportion result, where the higher-order thinking question types are the question types pre-selected from all question types;

[0015] Count the number of random questions and follow-up questions to obtain the total number of random questions and / or follow-up questions;

[0016] Based on the proportion result and the total number, obtain the classroom question analysis result;

[0017] Among them, the process of detecting random questions and follow-up questions includes:

[0018] Determine the question judgment time period according to the start and end times of the interrogative sentence;

[0019] Detect whether there is a student standing up in the video of the student panoramic video during the question judgment time period;

[0020] When there is a student standing up, judge that a teacher's question is detected;

[0021] Detect whether there is a student raising a hand in the preset second time period before the student stands up in the student panoramic video. When there is no student raising a hand, judge that the detected teacher's question type is a random question;

[0022] When a teacher's question is detected, obtain the sentence pattern judgment result of the teacher's audio data during the student's standing period. When there is an interrogative sentence in the sentence pattern judgment result of the teacher's audio data during the student's standing period, judge that a teacher's follow-up question is detected.

[0023] In one implementation, the process of determining the question judgment time period according to the start and end times of the interrogative sentence includes:

[0024] Use the start time of the interrogative sentence as the start time of the question judgment time period;

[0025] Use a time point after the end time of the interrogative sentence as the end time of the question judgment time period, where the difference between the end time of the interrogative sentence and the time point is a preset first time period.

[0026] In one implementation, it further includes:

[0027] Count the number of follow-up questions in the higher-order thinking question types, denoted as the third number;

[0028] Also obtain the classroom question analysis result based on the third number.

[0029] In one implementation, it further includes:

[0030] For each follow-up question, the number of interrogative sentences that appear in the sentence pattern judgment result of the teacher's audio data during the student's standing is also recorded, which is denoted as the continuous follow-up question number;

[0031] The classroom question-asking analysis result is also obtained according to each continuous follow-up question number.

[0032] In a third aspect, the present invention provides a device for judging the type of teacher's questions, including:

[0033] A receiving module, configured to obtain the content to be judged, where the content to be judged is obtained from the text data obtained by converting the teacher's voice into text;

[0034] A matching module, configured to match the content to be judged with a preset rule set to obtain the hit rules therein. Among them, there are multiple preset rules in the preset rule set, each preset rule has its corresponding question type, and each preset rule has its corresponding priority;

[0035] A judging module, configured to use the rule with the highest priority among the hit rules as the rule corresponding to the content to be judged, and extract the corresponding question type as the question type of the content to be judged.

[0036] A classroom question-asking analysis device, including:

[0037] An obtaining module, configured to obtain the sentence pattern judgment result of the text data obtained by converting the teacher's voice into text, where the sentence pattern judgment result is obtained through a semantic prediction model based on deep learning;

[0038] An extraction module, configured to extract the interrogative sentences in the text data according to the sentence pattern judgment result, and determine the extracted interrogative sentences as the content to be judged;

[0039] A judgment module, which is the device for judging the type of teacher's questions described above, to obtain the question type of each content to be judged;

[0040] A first statistics module, configured to calculate the proportion of high-order thinking question types in the question types to obtain a proportion result, where the high-order thinking question types are the question types pre-selected from all question types;

[0041] A second statistics module, configured to count the number of random questions and follow-up questions to obtain the total number of random questions and / or follow-up questions;

[0042] An analysis module, configured to obtain the classroom question-asking analysis result according to the proportion result and the total number;

[0043] Among them, the process of the second statistics module detecting random questions and follow-up questions includes:

[0044] Determine the question judgment time period according to the start and end times of the interrogative sentence;

[0045] Detect whether there is a student standing up in the video of the student panoramic video during the question judgment time period;

[0046] When there is a student standing up, determine that a teacher's question is detected;

[0047] Detect whether a student raises a hand within a preset second time period before the student stands up in the student panoramic video. When there is no student raising a hand, determine that the detected teacher's question type is a random question;

[0048] When a teacher's question is detected, obtain the sentence pattern judgment result of the teacher's audio data during the student's standing period. When there is an interrogative sentence in the sentence pattern judgment result of the teacher's audio data during the student's standing period, determine that a teacher's follow-up question is detected.

[0049] In a fourth aspect, the present invention provides a computer device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, the method of any of the above embodiments is implemented.

[0050] In a fifth aspect, the present invention provides a computer-readable storage medium, on which a computer program is stored. The program is characterized in that when it is executed by a processor, the method of any of the above embodiments is implemented.

[0051] In the present invention, the type of the teacher's question is judged through the question type corresponding to the preset rule and the priority between the preset rules. Based on the type of the teacher's question, combined with the Bloom's higher-order questioning cognition of follow-up questions, classroom question analysis is carried out. The teacher continuously asks inspiring questions during the follow-up process, indicating that the substantial communication between teachers and students is effective, so this can also be used as a judgment basis. The present invention is simple and practical, and fits the actual teaching scenario. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 is an overall flowchart of an embodiment of the first embodiment of the present invention.

[0053] Figure 2 is a schematic diagram of the Bloom model.

[0054] Figure 3 is an overall flowchart of an embodiment of the second embodiment of the present invention.

[0055] Figure 4 is a flowchart of the process of detecting random questions and follow-up questions in the second embodiment of the present invention.

[0056] Figure 5 is an overall structural diagram of an embodiment of the third embodiment of the present invention.

[0057] Figure 6 It is a schematic diagram of the overall structure of another implementation manner of Embodiment 3 of the present invention. Specific implementation manner

[0058] Hereinafter, exemplary embodiments of the present disclosure will be described in more detail with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided so that the present disclosure can be more thoroughly understood and the scope of the present disclosure can be completely conveyed to those skilled in the art.

[0059] It should be noted that the terms "first / second / ..." involved in the embodiments of the present invention are only used to distinguish similar objects and do not represent a specific order for the objects. It can be understood that "first / second / ..." can be interchanged with a specific order or sequence when permitted. It should be understood that the objects distinguished by "first / second / ..." can be interchanged under appropriate circumstances so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein.

[0060] Embodiment 1

[0061] Please refer to Figure 1 , Figure 1 which is a schematic flowchart of a method for judging the type of teacher's questions provided in Embodiment 1 of the present invention. The method includes step S110, step S120, and step S130. It should be noted that step S110, step S120, and step S130 are only reference numerals used to clearly explain the correspondence between the embodiment and the attached Figure 1 drawings and do not represent a limitation on the order of each step in this embodiment.

[0062] Step S110: Obtain the content to be determined, where the content to be determined is obtained from the text data obtained by converting the teacher's voice into text;

[0063] Step S120: Match the content to be determined with a preset rule set to obtain the rules that are hit. Among them, there are multiple preset rules in the preset rule set, each preset rule has its own corresponding question type, and each preset rule has its own corresponding priority;

[0064] Step S130: Use the rule with the highest priority among the hit rules as the rule corresponding to the content to be determined, and extract the corresponding question type as the question type of the content to be determined.

[0065] This method is suitable for use in a recording and broadcasting host. The recording and broadcasting host receives multiple video data streams, such as the video from the teacher's panoramic camera, the video from the teacher's tracking camera, the video from the student's panoramic camera, the video from the student's tracking camera, etc. In addition, it can also receive the video output from a computer, the video of the teaching blackboard writing, etc. In addition to receiving video data, the recording and broadcasting host also receives audio data, such as the audio from the teacher's microphone, the audio from the student's microphone / student microphone array, etc.

[0066] For the teacher's audio data received from the teacher's microphone, it is first processed by a deep learning model to obtain a speech-to-text result. This deep learning model can infer a sentence pattern judgment result when understanding the semantics and combining the context. Specifically, it can infer where to add what punctuation marks when understanding the semantics and combining the context. When there is a "?" in the speech-to-text input of the teacher's audio data, it can be determined that the sentence is an interrogative sentence. If it is an interrogative sentence, then the content in the interrogative sentence is retained, that is, the sentence content before the "?" is used as the content to be determined.

[0067] In step S110, the content to be determined obtained from the text data of the teacher's speech-to-text above is acquired.

[0068] As Figure 2 shown, Figure 2 is a schematic diagram of the Bloom model. The Bloom model divides the objectives into six levels, from low to high in sequence:

[0069] Remember: Remember specific knowledge, such as facts and basic concepts.

[0070] Understand: Understand what is learned and be able to retell it in one's own words.

[0071] Apply: Apply the learned knowledge to actual situations.

[0072] Analyze: Decompose the information into its components and understand its structure.

[0073] Evaluate: Judge and evaluate the information and make a reasonable conclusion based on evidence.

[0074] Create: Combine different elements into a new whole, emphasizing creative thinking.

[0075] Correspondingly, in this embodiment, the question types are also divided into six types: memory type, understanding type, application type, analysis type, evaluation type, and creativity type. For each type of question, several matching rules are preset. For example, one of the preset rules for the memory type can be: (can you recall | can you remember).*come [? ? ], which is a regular expression with the following meanings: ①Can you recall: means matching a string containing the five words "can you recall" or the five words "can you remember" in the text; ②.*: means matching any number of any characters (except line breaks), * means that the preceding character can appear any number of times, including zero times; ③come: means matching a string containing the two words "come" in the text; ④[? ? ]: means matching any one of the question mark? or Chinese question mark?; In summary, this regular expression is used to match a string that starts with "can you recall", can have any characters in the middle, is followed by "come", and ends with a question mark or Chinese question mark. For example, it can match teacher-asked questions like “Can you recall yesterday’s lesson?” or “Can you remember this story?”

[0076] In actual teaching scenarios, a teacher's words may trigger two preset rules at the same time, and these two preset rules may belong to different question types. What should be done at this time? In order to solve the above problem, this embodiment also adds a priority coefficient to each preset rule. For example, the priority coefficient of rule A is 42, the finite level coefficient of rule B is 52, and the finite level coefficient of rule C is 13. If the content to be determined triggers the three rules A, B, and C at the same time, it is considered that the content to be determined triggers the high-priority rule C, that is, as described in step S130, the rule with the highest priority among the hit rules is used as the rule corresponding to the content to be determined. At this time, if the question type corresponding to rule C is analytical, the content to be determined is considered to be of the analytical type.

[0077] Preferably, the preset rules include: key character strings, logical relationships between key character strings, and a sequence.

[0078] As shown in the preset rules above, the preset rules may include key strings and the logical relationship between these key strings, such as "can you recall", "can you remember", "come [? ?]", etc. These are key strings, and the regular expressions between these key strings are used to describe the logic and sequence between them.

[0079] In this method, the type of question the teacher is asking is determined by the type of question corresponding to the preset rules and the priority between the preset rules. This method is simple and practical, and fits the actual teaching scenario.

[0080] Example 2

[0081] Please refer to Figure 3 , Figure 3 , which is a schematic flowchart of a classroom question - asking analysis method provided in the second embodiment of the present invention. This method includes steps S210, S220, S230, S240, S250, and S260. It should be noted that steps S210, S220, S230, S240, S250, and S260 are only reference signs used to clearly explain the corresponding relationship between the embodiment and the attached Figure 3 drawings, and do not represent the order limitation of each step in this embodiment.

[0082] Step S210: Obtain the sentence - pattern judgment result of the text data obtained by converting the teacher's voice into text, where the sentence - pattern judgment result is obtained through a semantic prediction model based on deep learning;

[0083] Step S220: Extract the interrogative sentences from the text data according to the sentence - pattern judgment result, and determine the extracted interrogative sentences as the content to be judged;

[0084] Step S230: Use the judgment method of the teacher's question type in Embodiment 1 to obtain the question type of each content to be judged;

[0085] Step S240: Calculate the proportion of the high - order thinking question type in the question types, and obtain the proportion result, where the high - order thinking question type is the question type pre - selected from all question types;

[0086] Step S250: Count the number of times of random questions and follow - up questions to obtain the total number of random questions and / or follow - up questions;

[0087] Step S260: Obtain the classroom question - asking analysis result according to the proportion result and the total number;

[0088] Among them, please refer to Figure 4 , Figure 4 , which is a schematic flowchart of the process of detecting random questions and follow - up questions of the present invention. The process of detecting random questions and follow - up questions includes: steps S310, S320, S330, S340, and S350.

[0089] Step S310: Determine the question - asking judgment time period according to the start and end times of the interrogative sentence;

[0090] Step S320: Detect whether there is a student standing up in the video of the student panoramic video during the question - asking judgment time period;

[0091] Step S330: When there is a student standing up, judge that a teacher's question has been detected;

[0092] Step S340: Detect whether a student raises a hand within a preset second duration before the student stands up. When no student raises a hand, determine that the detected teacher's question type is a random question.

[0093] Step S350: When it is detected that the teacher asks a question, obtain the sentence pattern judgment result of the teacher's audio data during the student's standing period. When the sentence pattern judgment result of the teacher's audio data during the student's standing period is an interrogative sentence, determine that the teacher's follow-up question is detected.

[0094] As Figure 2 shown, there are six types of question types. If the teacher asks more high-order thinking question types in classroom questions, it can better reflect the substantial communication between the teacher and the students and better reflect the effect of classroom interaction. Therefore, in the second embodiment, each question asked by the teacher is analyzed, and the quality of the questions is analyzed through the proportion of high-order thinking question types.

[0095] Specifically, first, as described in the first embodiment, determine the interrogative sentences in the text data and extract them. Then, use the judgment method of the teacher's question type in the first embodiment to judge the extracted content and obtain the question type results of each interrogative sentence. Among them, application type, analysis type, evaluation type, and creation type are defined as high-order thinking question types, and calculate their proportion, that is, the result of (application type + analysis type + evaluation type + creation type) / (memory type + understanding type + application type + analysis type + evaluation type + creation type). Finally, obtain a first analysis result according to the proportion. For example, if the proportion of high-order dialogue is less than 1%, get 1 point; if the proportion of high-order dialogue is between 1% and 10%, get 2 points; if the proportion of high-order dialogue is between 11% and 30%, get 3 points; if the proportion of high-order dialogue exceeds 31%, get 4 points.

[0096] In addition to the proportion of high-order thinking question types proposed by the teacher, in this embodiment, the number of random questions and follow-up questions asked by the teacher is also used to judge whether there is substantial communication between the teacher and the students, and to evaluate the effect of classroom interaction accordingly.

[0097] Meanwhile, in the second embodiment, a process for detecting random questions and follow-up questions is also provided.

[0098] Specifically, when the teacher asks a question, the teacher naturally asks with an interrogative sentence. However, since the teacher saying an interrogative sentence does not necessarily mean that the teacher is asking a question, after capturing the situation where the sentence pattern judgment result is an interrogative sentence, further verification is required through certain means. To improve the efficiency of further verification, verification can be performed only in the time period near when the teacher says the interrogative sentence. Therefore, in step S310, according to the start and end times of the interrogative sentence, determine the question judgment time period.

[0099] In one implementation, the process of determining the question-asking judgment time period according to the start and end times of the interrogative sentence in step S310 includes: step S311 and step S312.

[0100] Step S311, using the start time of the interrogative sentence as the start time of the question-asking judgment time period;

[0101] Step S312, using a time point after the end time of the interrogative sentence as the end time of the question-asking judgment time period, where the difference between the end time of the interrogative sentence and the time point is a preset first duration.

[0102] In this implementation, it starts when the teacher utters the interrogative sentence, so the start time of the interrogative sentence is used as the start time of the question-asking judgment time period. Additionally, in many cases, after the teacher finishes an interrogative sentence, everyone will have a reaction or response. Therefore, in step S312, after the teacher finishes the interrogative sentence, a preset first duration, such as 1 minute, is left for waiting for everyone's reaction or response.

[0103] Step S320 is the specific process of the further verification mentioned above. In this method, the video of the student panoramic camera is received. At this time, the received student panoramic video is analyzed to detect whether a student stands up during the question-asking judgment time period. In a classroom scenario, after the teacher asks a question, the student generally stands up to answer the teacher's question. Therefore, step S320 uses whether the action of a student standing up occurs as the basis for further verification. It should be noted here that what is detected is the action of standing up, that is, the process of changing from sitting to standing, not the action of originally standing and always maintaining the standing posture.

[0104] Finally, in step S330, when it is detected that a student stands up during the question-asking judgment time period, it can be determined that a teacher's question-asking event has occurred, that is, it is detected that the teacher asks a question at this time.

[0105] When the teacher asks a question, it may be an ordinary question, or it may be a random question or a follow-up question. Therefore, it is necessary to further judge the type of the teacher's question and screen out the cases of random questions and follow-up questions.

[0106] Step S340 is to further refine the question type and detect the process of a random question. Generally speaking, when the teacher asks a random question, the student does not need to raise his hand. Therefore, in step S340, it is detected whether the student raises his hand before standing up. If not, the question type is judged as a random question; otherwise, the question type is judged as an ordinary question.

[0107] Step S350 is a detection process for the question type of follow-up questions. A follow-up question means that the teacher continues to ask questions after the student answers the first question, which is another way of interacting with the student. In this embodiment, by retrieving the sentence pattern judgment result during the student's standing period, it is judged whether the teacher has asked a question based on whether there is an interrogative sentence. If the teacher has asked a question, it is judged that a teacher's follow-up question has been detected.

[0108] This process of detecting random questions and follow-up questions can simply, conveniently and effectively detect the teacher's questions and the types of questions, filling the technical gap in the judgment method of the teacher's question types and providing a technical basis for the subsequent teacher-student interaction analysis.

[0109] Continuing with the previous steps S250 and S260, step S260 obtains the classroom question analysis result according to the total number of random questions and / or follow-up questions. For example, if the number of random questions and / or follow-up questions is 0, 1 point is obtained; if the number of random questions and / or follow-up questions is greater than 2, 2 points are obtained; if the number of random questions and / or follow-up questions is greater than 3, 3 points are obtained; if the number of random questions and / or follow-up questions is greater than 4, 4 points are obtained.

[0110] It should be noted here that the meaning of random questions and / or follow-up questions can be to only count the random questions, or only count the follow-up questions, or count the total of both. Those skilled in the art can adjust the required scheme according to actual needs.

[0111] In this method, it is judged whether there is a substantial communication between the teacher and the student based on the proportion of the high-order thinking question types proposed by the teacher and the number of random questions and follow-up questions, and the effect of classroom interaction is evaluated accordingly. The entire evaluation process conforms to the actual teaching scenario and is simple and reasonable.

[0112] In one implementation manner, it further includes: steps S271 and S272.

[0113] Step S271, count the number of follow-up questions among the high-order thinking question types, and record it as the third number;

[0114] Step S272, also obtain the classroom question analysis result according to the third number.

[0115] This implementation manner combines Bloom's higher-order questioning cognition of follow-up questions for classroom question analysis. The teacher continuously asks inspiring questions during the follow-up process, indicating that the substantial communication between the teacher and the student is effective, so this can also be used as a judgment basis.

[0116] In one implementation manner, it further includes: steps S281 and S282.

[0117] Step S281, for each follow-up question, also record the number of times an interrogative sentence appears in the sentence pattern judgment result of the teacher's audio data during the student's standing period, and record it as the continuous follow-up number;

[0118] In step S282, the classroom questioning analysis result is also obtained according to the number of consecutive questioning times for each case.

[0119] For more suitable evaluation in the teaching scenario, in this embodiment, the number of consecutive questionings during each questioning is also counted. Through the statistics of the number of consecutive questionings, the teacher-student interaction situation in the questioning process in teaching can be further reflected.

[0120] It should be noted here that the analysis result in step S282 can be obtained through the average value of the number of consecutive questionings, or different scores are assigned to different number intervals, and finally the analysis result is obtained through the weighting of these scores. In this embodiment, the specific process of obtaining the analysis result is not limited.

[0121] Embodiment III

[0122] Corresponding to the method of Embodiment I, as Figure 5 shown, the present invention also provides a device 5 for detecting teacher's questions, including: a receiving module 510, a matching module 520, and a determination module 530.

[0123] The receiving module 510 is configured to obtain the content to be determined, where the content to be determined is obtained from the text data obtained by converting the teacher's voice into text;

[0124] The matching module 520 is configured to match the content to be determined with a preset rule set to obtain the hit rules therein. Among the preset rule sets, there are multiple preset rules, each preset rule has its corresponding question type, and each preset rule has its corresponding priority;

[0125] The determination module 530 is configured to use the rule with the highest priority among the hit rules as the rule corresponding to the content to be determined, and extract the corresponding question type as the question type of the content to be determined.

[0126] In one embodiment, the preset rules include: key strings, the logical relationship between key strings, and the sequence.

[0127] In this device, the type of the teacher's question is judged through the question type corresponding to the preset rule and the priority among the preset rules. This device is simple and practical, and conforms to the actual teaching scenario.

[0128] Corresponding to the method of Embodiment II, as Figure 6 shown, the present invention also provides a classroom questioning analysis device 6, including: an acquisition module 610, an extraction module 620, a judgment module 630, a first statistics module 640, a second statistics module 650, and an analysis module 660.

[0129] An acquisition module 610, configured to acquire a sentence pattern judgment result of text data obtained by converting a teacher's speech into text, where the sentence pattern judgment result is obtained by a semantic prediction model based on deep learning;

[0130] An extraction module 620, configured to extract interrogative sentences from the text data according to the sentence pattern judgment result, and determine the extracted interrogative sentences as content to be determined;

[0131] A judgment module 630, configured to obtain the question type of each content to be determined by using the foregoing judgment device for the teacher's question type;

[0132] A first statistics module 640, configured to calculate the proportion of high-order thinking question types in the question types to obtain a proportion result, where the high-order thinking question type is a question type pre-selected from all question types;

[0133] A second statistics module 650, configured to count the number of random questions and the number of follow-up questions to obtain the total number of random questions and / or follow-up questions;

[0134] An analysis module 660, configured to obtain a classroom question analysis result according to the proportion result and the total number;

[0135] Among them, the process of the second statistics module detecting random questions and follow-up questions includes:

[0136] Determine a question judgment time period according to the start time and end time of the interrogative sentence;

[0137] Detect whether there is a student standing up in the video of the student panoramic video during the question judgment time period;

[0138] When there is a student standing up, determine that a teacher's question is detected;

[0139] Detect whether there is a student raising a hand in a preset second time period before the student stands up in the student panoramic video. When there is no student raising a hand, determine that the detected teacher's question type is a random question;

[0140] When it is detected that there is a teacher's question, obtain a sentence pattern judgment result of the teacher's audio data during the period when the student stands. When there is an interrogative sentence in the sentence pattern judgment result of the teacher's audio data during the period when the student stands, determine that a teacher's follow-up question is detected.

[0141] In one implementation manner, the process of determining a question judgment time period according to the start time and end time of the interrogative sentence includes:

[0142] Use the start time of the interrogative sentence as the start time of the question judgment time period;

[0143] Taking a time point after the end time of the interrogative sentence as the end time of the question judgment time period, wherein the end time of the interrogative sentence and the time point differ by a preset first duration.

[0144] In one implementation, it further includes: a third statistics module;

[0145] The third statistics module is used to count the number of times of follow-up questions in the high-order thinking question types, denoted as the third number;

[0146] The analysis module is further used to obtain the classroom question analysis result according to the third number.

[0147] In one implementation, it further includes: a third statistics module;

[0148] The fourth statistics module is used to record, for each follow-up question, the number of times an interrogative sentence appears in the sentence pattern judgment result of the teacher's audio data during the student's standing period, denoted as the consecutive follow-up number;

[0149] The analysis module is further used to obtain the classroom question analysis result according to each consecutive follow-up number.

[0150] In this device, it is judged whether there is substantial communication between the teacher and the students through the proportion of the high-order thinking question types proposed by the teacher and the number of random questions and follow-up questions, and the effect of classroom interaction is evaluated based on this. The entire evaluation process conforms to the actual teaching scenario and is simple and reasonable.

[0151] Example 4

[0152] The embodiment of the present invention also provides a storage medium, on which computer instructions are stored, and when the instructions are executed by a processor, the judgment method of the teacher's question type and / or the classroom question analysis method in any of the above embodiments are implemented.

[0153] Those skilled in the art can understand that all or part of the steps of implementing the above method embodiments can be completed by hardware related to program instructions. The foregoing program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps including the above method embodiments; and the foregoing storage medium includes: various media such as mobile storage devices, random access memory (RAM), read-only memory (ROM), magnetic disks, or optical discs that can store program codes.

[0154] Alternatively, if the above integrated units of the present invention are implemented in the form of software functional modules and sold or used as independent products, they can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiments of the present invention, in essence or the part that contributes to the related technology, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a terminal, or a network device, etc.) to execute all or part of the methods of the various embodiments of the present invention. The aforementioned storage medium includes: various media such as removable storage devices, RAM, ROM, magnetic disks, or optical discs that can store program codes.

[0155] Corresponding to the above computer storage medium, in one embodiment, a computer device is further provided. The computer device includes a memory, an encoder, and a computer program stored on the memory and executable on the encoder. When the encoder executes the program, it implements the judgment method of any one of the teacher's question types and / or the classroom question analysis method in the above embodiments.

[0156] The above computer device determines the type of the teacher's question through the question types corresponding to the preset rules and the priorities among the preset rules. Based on the type of the teacher's question, classroom question analysis is carried out in combination with Bloom's higher-order questioning cognition of follow-up questions. The teacher continuously asks inspiring questions during the process of follow-up questions, indicating that the substantial communication between teachers and students is effective. Therefore, this can also be used as a judgment basis. This computer device is simple and practical, and conforms to the actual teaching scenario.

[0157] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope described in this specification.

[0158] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, rather than limitations on the implementation manners of the present invention. For those of ordinary skill in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the implementation manners here. Any modifications, equivalent replacements, and improvements made within the spirit and principle of the present invention shall be included in the protection scope of the claims of the present invention.

Claims

1. A method for judging the type of teacher's questions, characterized in that, Including: Obtain the content to be judged, where the content to be judged is obtained from the text data obtained by converting the teacher's speech into text; Match the content to be judged with a preset rule set to obtain the hit rules. Among the preset rule sets, there are multiple preset rules, each preset rule has its corresponding question type, and each preset rule has its corresponding priority; Use the rule with the highest priority among the hit rules as the rule corresponding to the content to be judged, and extract the corresponding question type as the question type of the content to be judged.

2. The method for judging the teacher's question type according to claim 1, wherein The preset rules include: key strings, the logical relationship between key strings, and the sequence.

3. A method for analyzing classroom questions, characterized in that, Also including: Obtain the sentence pattern judgment result of the text data obtained by converting the teacher's speech into text, where the sentence pattern judgment result is obtained through a semantic prediction model based on deep learning; Extract the interrogative sentences in the text data according to the sentence pattern judgment result, and determine the extracted interrogative sentences as the content to be judged; Use the method for judging the teacher's question type described in claim 1 or 2 to obtain the question types of each content to be judged; Calculate the proportion of higher-order thinking question types in the question types to obtain a proportion result, where the higher-order thinking question types are the question types pre-selected from all question types; Count the number of times of random questions and follow-up questions to obtain the total number of random questions and / or follow-up questions; Obtain the classroom question analysis result according to the proportion result and the total number; Among them, the process of detecting random questions and follow-up questions includes: Determine the question judgment time period according to the start and end times of the interrogative sentence; Detect whether there is a student standing up in the video of the student panoramic video during the question judgment time period; When a student stands up, it is judged that a teacher's question is detected; Detect whether there is a student raising a hand in a preset second time period before the student stands up. When there is no student raising a hand, it is judged that the detected teacher's question type is a random question; When a teacher's question is detected, obtain the sentence pattern judgment result of the teacher's audio data during the student's standing period. When there is an interrogative sentence in the sentence pattern judgment result of the teacher's audio data during the student's standing period, it is judged that a teacher's follow-up question is detected.

4. The classroom question analysis method according to claim 3, characterized in that The process of determining the question judgment time period according to the start and end times of the interrogative sentence includes: Use the start time of the interrogative sentence as the start time of the question judgment time period; Use a time point after the end time of the interrogative sentence as the end time of the question judgment time period, where the difference between the end time of the interrogative sentence and the time point is a preset first time period.

5. The classroom question analysis method according to claim 3 or 4, characterized in that, Also including: Count the number of times of follow-up questions in the higher-order thinking question types, denoted as the third number; Also obtain the classroom question analysis result according to the third number.

6. The classroom question analysis method according to claim 3 or 4, characterized in that, Also including: For each follow-up question, also record the number of times an interrogative sentence appears in the sentence pattern judgment result of the teacher's audio data during the student's standing period, denoted as the continuous follow-up number; Also obtain the classroom question analysis result according to each continuous follow-up number.

7. A judging device for the type of teacher's questions, characterized in that, Including: A receiving module for obtaining the content to be judged, where the content to be judged is obtained from the text data obtained by converting the teacher's speech into text; A matching module, configured to match the content to be determined with a preset rule set to obtain the rules that are hit. Among them, there are multiple preset rules in the preset rule set, each preset rule has its corresponding question type, and each preset rule has its own corresponding priority; A determination module, configured to use the rule with the highest priority among the hit rules as the rule corresponding to the content to be determined, and extract the corresponding question type as the question type of the content to be determined.

8. A classroom question analysis device, characterized in that, including: An acquisition module, configured to obtain the sentence pattern judgment result of the text data obtained by converting the teacher's voice into text, where the sentence pattern judgment result is obtained by a semantic prediction model based on deep learning; An extraction module, configured to extract interrogative sentences from the text data according to the sentence pattern judgment result, and determine the extracted interrogative sentences as the content to be determined; A judgment module, configured to use the judgment device for the teacher's question type described in claim 7 to obtain the question type of each content to be determined; A first statistics module, configured to calculate the proportion of high-order thinking question types in the question types to obtain a proportion result, where the high-order thinking question type is a question type pre-selected from all question types; A second statistics module, configured to count the number of random questions and follow-up questions to obtain the total number of random questions and / or follow-up questions; An analysis module, configured to obtain a classroom question analysis result according to the proportion result and the total number; Among them, the process of the second statistics module detecting random questions and follow-up questions includes: Determining a question judgment time period according to the start and end times of the interrogative sentence; Detecting whether there is a student standing up in the video of the student panoramic video during the question judgment time period; When there is a student standing up, determining that a teacher's question is detected; Detecting whether there is a student raising a hand in a preset second time period before the student stands up. When there is no student raising a hand, determining that the detected teacher's question type is a random question; When a teacher's question is detected, obtaining the sentence pattern judgment result of the teacher's audio data during the student's standing period. When there is an interrogative sentence in the sentence pattern judgment result of the teacher's audio data during the student's standing period, determining that a teacher's follow-up question is detected.

9. A computer device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the program, the method described in any one of claims 1-6 is implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, the method described in any one of claims 1-6 is implemented.

Images