An intelligent test paper compiling method and device

Abstract

The application discloses an intelligent test paper compiling method and device, and belongs to the technical field of image or video recognition or understanding. The intelligent test paper compiling method comprises the following steps: collecting original data of various types of questions, and performing standardization processing on the collected data to form a standard question bank; wherein the standardization processing comprises adding applicable teaching materials, applicable grades, question types, first-level knowledge point labels, second-level knowledge point labels, difficulty coefficients and score information to the original data; performing pre-de-duplication processing on the questions in the standard question bank, then dividing the standard question bank into a plurality of first-level question banks according to the first-level knowledge point labels, and dividing the first-level question banks into a plurality of second-level question banks according to the second-level knowledge point labels; and screening questions from the standard question bank according to set test paper compiling information. When the repeated questions belong to the second-level question banks related to core knowledge points, the repeated questions are replaced by another question in the same second-level question bank, so that the accuracy of test paper compiling is ensured again.

Description

Technical Field

[0001] This invention belongs to the field of image or video recognition or understanding technology; specifically, it relates to an intelligent document generation method and apparatus. Background Technology

[0002] In the field of education, test paper compilation is a crucial and complex task. The traditional manual test paper compilation model has significant limitations: teachers need to spend a lot of time and energy to complete the selection of test questions, allocation of marks, and balance of difficulty. After that, they also need to carry out test paper grading, data analysis and archiving management. The overall efficiency is low and easily affected by subjective factors, which may lead to unreasonable test question structure or biased assessment results.

[0003] Chinese Patent CN114005116A discloses a method, apparatus, and storage medium for assembling exam papers based on exam paper information. The method includes: acquiring exam paper attribute information of the exam paper to be assembled; retrieving a set of similar exam papers from a database based on the exam paper attribute information; analyzing the set of similar exam papers to obtain exam paper structure information, question knowledge point information, and question difficulty information of the exam paper to be assembled; retrieving similar questions from a question bank using the exam paper structure information, question knowledge point information, and question difficulty information as filtering conditions; and assembling the retrieved similar questions into an exam paper. This invention's exam paper assembly method achieves automatic exam paper assembly according to assembly requirements, meeting the requirements for exam paper assembly and greatly improving the efficiency of exam paper assembly.

[0004] However, the aforementioned patent documents cannot achieve accurate deduplication during the document assembly process with relatively low computing power. Summary of the Invention

[0005] The purpose of this disclosure is to provide an intelligent paper-generating method and apparatus that can solve at least some of the above-mentioned technical problems.

[0006] A first aspect of this disclosure provides an intelligent document generation method, the intelligent document generation method comprising:

[0007] Raw data of various types of questions are collected, and the collected data is standardized to form a standard question bank. The standardization process includes adding information such as applicable textbooks, applicable grades, question types, first-level knowledge point tags, second-level knowledge point tags, difficulty coefficients, and score values ​​to the raw data.

[0008] The questions in the standard question bank are pre-deduplicated, and then the standard question bank is divided into multiple first-level question banks according to the first-level knowledge point tags, and the first-level question banks are divided into multiple second-level question banks according to the second-level knowledge point tags.

[0009] Based on the set test paper information, questions are selected from the pre-deduplicated standard question bank. The specific process is as follows:

[0010] Based on the total score and question type requirements of the test paper, the number of each question type and the score range for each question should be reasonably allocated.

[0011] Based on the preset difficulty distribution ratio, select questions with the corresponding difficulty level;

[0012] At least one question is selected from each of the primary question banks associated with the test paper information, with priority given to selecting from the secondary question banks related to the core knowledge points;

[0013] After the initial test paper compilation is completed, the selected questions undergo a final deduplication process. During this final deduplication process, if duplicate questions are detected, it is first determined whether the duplicate questions belong to a secondary question bank related to the core knowledge points. If the duplicate questions belong to a secondary question bank related to the core knowledge points, they are replaced with another question from the same secondary question bank. If the duplicate questions do not belong to a secondary question bank related to the core knowledge points, they are replaced with another question from the same primary question bank.

[0014] Optionally, the primary knowledge point tags include multiple associated secondary knowledge point tags.

[0015] Optionally, the final deduplication process includes a primary detection method, which detects whether the question type and secondary knowledge point tags of the selected questions are duplicated. When there are two questions in the selected questions with the same question type and secondary knowledge point tags, the primary detection method determines that the selected questions contain duplicate knowledge points.

[0016] Optionally, the final deduplication process further includes a secondary detection method, which determines whether the similarity between the answers and solution steps of the selected questions exceeds a first preset threshold. If the similarity between the answers and solution steps of two selected questions exceeds the first preset threshold, the secondary detection method determines that the selected questions contain duplicate knowledge points. Specifically, if the primary detection method determines that the selected questions contain duplicate knowledge points, the process proceeds to the secondary detection method; if the primary detection method determines that the selected questions do not contain duplicate knowledge points, the process does not proceed to the secondary detection method.

[0017] Optionally, the final deduplication process further includes a three-level detection method, which determines whether the semantic similarity of the selected questions exceeds a second preset threshold. If the semantic similarity between two selected questions exceeds the second preset threshold, the three-level detection method determines that the selected questions contain duplicate knowledge points. Specifically, if the two-level detection method determines that the selected questions contain duplicate knowledge points, the process proceeds to the three-level detection method. If the two-level detection method determines that the selected questions do not contain duplicate knowledge points, the process does not proceed to the three-level detection method.

[0018] Optionally, the three-level detection method uses BERT to determine the semantic similarity of the selection questions. The specific steps for determining the semantic similarity of the selection questions using BERT include:

[0019] Clean the selection questions: remove irrelevant content and ensure a uniform format;

[0020] Word segmentation encoding: The sentence in the selection question is broken down into words, and each word is encoded;

[0021] Generating sentence fingerprints: Let BERT process the encoded text to obtain the feature vectors of all words, and then average the vectors of all words to obtain the sentence fingerprints for selecting questions;

[0022] Similarity comparison: Calculate the cosine similarity between the sentence fingerprints of two screening questions.

[0023] Optionally, the final deduplication process involves determining whether the semantic similarity of the selected questions exceeds a third preset threshold.

[0024] A second aspect of the present disclosure provides an electronic device, characterized in that,

[0025] Including the processor and memory;

[0026] The memory is used to store programs;

[0027] The processor executes the program to implement the method described above.

[0028] A third aspect of the present disclosure provides a computer-readable storage medium, characterized in that,

[0029] The storage medium stores a program, which is executed by a processor to implement the method described above.

[0030] A fourth aspect of this disclosure provides an intelligent document generation device, which includes:

[0031] The data acquisition module is used to collect raw data of various types of questions and to standardize the collected data to form a standard question bank. The standardization process includes adding information such as applicable textbook, applicable grade, question type, first-level knowledge point tag, second-level knowledge point tag, difficulty coefficient, and score to the raw data.

[0032] The preprocessing module is used to pre-deduplicate the questions in the standard question bank, and then divide the standard question bank into multiple first-level question banks according to the first-level knowledge point tags, and divide the first-level question banks into multiple second-level question banks according to the second-level knowledge point tags.

[0033] The test paper creation rule setting module provides a human-computer interaction interface for users to input test paper creation information, and parses and stores the input information;

[0034] The question filtering and matching module filters out questions that meet the set conditions from the question bank based on the set test paper information.

[0035] The test paper generation and adjustment module is used to reasonably allocate the number of each question type and the score range of a single question according to the total score of the test paper and the question type configuration requirements;

[0036] Used to filter questions with corresponding difficulty levels based on a preset difficulty distribution ratio;

[0037] Used to select at least one question from each of the primary question banks associated with the test paper information, and to prioritize selection from the secondary question banks related to core knowledge points;

[0038] This is used to perform final deduplication on the selected questions after the initial test paper compilation is completed. During the final deduplication process, if it is determined that the selected questions are duplicated, it is first determined whether the duplicate questions belong to the secondary question bank related to the core knowledge points. If it is determined that the duplicate questions belong to the secondary question bank related to the core knowledge points, then the duplicate questions are replaced with another question from the same secondary question bank; if it is determined that the duplicate questions do not belong to the secondary question bank related to the core knowledge points, then the duplicate questions are replaced with another question from the same primary question bank.

[0039] The technical effects of this embodiment are as follows: By collecting and standardizing question data, the standard question bank is divided into a primary question bank and a secondary question bank. The primary question bank has a broad knowledge base, while the secondary question bank has precise knowledge points. In the process of test paper assembly, questions are first selected from the primary question bank associated with the test paper information to ensure the comprehensiveness of the test points. Then, the questions in the test paper are finally deduplicated. In the final deduplication process, it is prioritized to determine whether the duplicate questions belong to the secondary question bank related to the core knowledge points. When the duplicate questions belong to the secondary question bank related to the core knowledge points, the duplicate questions are replaced with another question from the same secondary question bank, thereby ensuring the accuracy of test paper assembly again. Attached Figure Description

[0040] Figure 1 This is a flowchart illustrating an intelligent paper-generating method according to an embodiment of the present disclosure;

[0041] Figure 2 This is a schematic diagram of a process for selecting questions from a standard question bank according to an embodiment of the present disclosure. Detailed Implementation

[0042] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The manner described in the following exemplary embodiments does not represent all manner consistent with this disclosure. Rather, they are merely examples of apparatuses consistent with some aspects of this disclosure as detailed in the appended claims.

[0043] The terminology used in this disclosure is for the purpose of describing particular embodiments only and is not intended to limit the disclosure. Unless otherwise defined, the technical or scientific terms used in this disclosure should be understood in their ordinary sense by one of ordinary skill in the art to which this disclosure pertains. The terms “first,” “second,” and similar terms used in this disclosure and the claims do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Similarly, the terms “a” or “one,” and similar terms do not indicate a quantity limitation, but rather indicate the presence of at least one, which will be separately stated if referring only to “a.” “A plurality” or “several” means two or more. Unless otherwise indicated, the terms “front,” “rear,” “lower,” and / or “upper,” and similar terms are for ease of description only and are not limited to a location or spatial orientation. The terms “comprising,” “including,” or “including,” and similar terms mean that the elements or objects preceding “comprising” or “including” encompass the elements or objects listed following “comprising” or “including” and their equivalents, but do not exclude other elements or objects. The terms “connection” or “link” are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. The singular forms “a,” “the,” and “the” used in this disclosure and the appended claims are also intended to include the plural forms, unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any and all possible combinations of one or more of the associated listed items.

[0044] like Figure 1 and Figure 2 As shown, the first aspect of this disclosure provides an intelligent paper-generating method, which includes:

[0045] Raw data from various question types is collected and standardized to form a standard question bank. This standardization process includes adding information such as applicable textbook, applicable grade level, question type, primary knowledge point tags, secondary knowledge point tags, difficulty level, and score to the raw data. Standardization eliminates differences between data from different sources, ensuring data consistency and comparability. For example, questions from different textbooks may use different wording or knowledge point tags; standardization unifies these questions into a common framework, facilitating subsequent screening and management.

[0046] The standard question bank undergoes pre-duplicate processing. Then, based on primary knowledge point tags, the standard question bank is divided into multiple primary question banks, and based on secondary knowledge point tags, the primary question banks are divided into multiple secondary question banks. Pre-duplicate processing quickly identifies redundant questions by detecting the repetition of question types and knowledge point tags. For example, if two questions are identical in type and knowledge point tag, they can be identified as duplicate questions. This initial deduplication significantly reduces redundant data in the question bank, improves the efficiency of subsequent screening, thereby increasing the purity of the question bank, reducing the waste of computational resources, and ensuring the efficiency of the subsequent test paper assembly process. Sub-bank management categorizes questions according to the hierarchical structure of knowledge points, making question screening more efficient and accurate. For example, primary knowledge point tags can be "solid geometry," "plane geometry," etc., and secondary knowledge point tags can be "triangle," "parallelogram," etc. Through this hierarchical management, test paper assemblers can quickly locate questions for specific knowledge points.

[0047] Based on the set test paper information, questions are selected from the pre-deduplicated standard question bank. The specific process is as follows:

[0048] Based on the total score and question type requirements of the exam paper, the number of each question type and the score range for each question are allocated reasonably. The exam paper selection process involves analyzing the total score and question type requirements to calculate the number of each question type and its score range. For example, if the total score is 100 points, multiple-choice questions account for 50 points, fill-in-the-blank questions for 30 points, and problem-solving questions for 20 points, then questions meeting these criteria are selected from the question bank. This reasonable allocation ensures a well-structured and comprehensive exam paper.

[0049] Based on the preset difficulty distribution ratio, select questions with the corresponding difficulty level;

[0050] At least one question is selected from each primary question bank associated with the test paper information, with priority given to selecting questions from secondary question banks related to core knowledge points. Core knowledge points are typically the focus of teaching and examinations; prioritizing questions on these knowledge points ensures the relevance and effectiveness of the test paper. For example, if the core knowledge point is "parallelogram," questions are prioritized from secondary question banks related to "parallelogram," thereby ensuring the test paper highlights key points, covers core knowledge points, and improves the quality of the test paper and teaching effectiveness.

[0051] After the initial test paper compilation is completed, a final deduplication process is performed on the selected questions. During this final deduplication, if duplicate questions are detected, the system first checks whether the duplicate question belongs to a secondary question bank related to the core knowledge points. If it does, the duplicate question is replaced with another question from the same secondary question bank. If it does not, the duplicate question is replaced with another question from the same primary question bank. This deduplication process is repeated after the initial test paper compilation to ensure there are no duplicate questions, prioritizing questions related to core knowledge points to avoid affecting the overall quality of the test paper.

[0052] In summary, by collecting and standardizing question data, the standard question bank is divided into a primary question bank and a secondary question bank. The primary question bank has a broad knowledge base, while the secondary question bank has precise knowledge points. During the test paper assembly process, questions are first selected from the primary question bank that is related to the test paper information to ensure the comprehensiveness of the test points. Then, the questions in the test paper are finally deduplicated. In the final deduplication process, priority is given to determining whether duplicate questions belong to the secondary question bank related to the core knowledge points. When a duplicate question belongs to the secondary question bank related to the core knowledge points, the duplicate question is replaced with another question from the same secondary question bank, thereby ensuring the accuracy of the test paper assembly.

[0053] As an optional implementation, the primary knowledge point tags include multiple associated secondary knowledge point tags. The inclusion of multiple secondary knowledge point tags within primary knowledge point tags clarifies the hierarchical relationship between knowledge points, facilitating accurate matching of user needs during question selection.

[0054] As an optional implementation, the final deduplication process includes a primary detection method, which detects whether the question type and secondary knowledge point tags of the selected questions are duplicated. When there are two questions in the selected questions with the same question type and secondary knowledge point tags, the primary detection method determines that the selected questions contain duplicate knowledge points.

[0055] The final deduplication process also includes a secondary detection method, which determines whether the similarity between the answers and solution steps of the selected questions exceeds a first preset threshold. If the similarity between the answers and solution steps of two selected questions exceeds the first preset threshold, the secondary detection method determines that the selected questions contain duplicate knowledge points. Specifically, if the primary detection method determines that the selected questions contain duplicate knowledge points, the process proceeds to the secondary detection method; if the primary detection method determines that the selected questions do not contain duplicate knowledge points, the process does not proceed to the secondary detection method.

[0056] The final deduplication process also includes a three-level detection method. The three-level detection method determines whether the semantic similarity of the selected questions exceeds a second set threshold. If the semantic similarity between two selected questions exceeds the second set threshold, the three-level detection method determines that the selected questions contain duplicate knowledge points. Specifically, if the two-level detection method determines that the selected questions contain duplicate knowledge points, the process proceeds to the three-level detection method. If the two-level detection method determines that the selected questions do not contain duplicate knowledge points, the process does not proceed to the three-level detection method.

[0057] The final deduplication process prioritizes checking for duplicate question types and secondary knowledge point tags. Next, it checks if the similarity between the answers and solution steps of the selected questions exceeds a first threshold. Finally, it checks if the semantic similarity of the selected questions exceeds a second threshold. This order of checks allows the processor to remove most of the original data using only a small amount of computing power through primary detection. Then, secondary and tertiary detection methods, which require more computing power, process the smaller portion of data filtered out by primary detection. This allows a typical processor to deduplicatize the entire original data, thus reducing costs.

[0058] As an optional implementation, the three-level detection method uses BERT to determine the semantic similarity of the selection questions. The specific steps for determining the semantic similarity of the selection questions using BERT include:

[0059] The selection process involves cleaning the questions: removing irrelevant content and ensuring a consistent format. Questions may contain irrelevant information (such as redundant symbols or inconsistent formatting), which can negatively impact subsequent semantic analysis. Data cleaning eliminates these interfering factors, improving the accuracy of semantic similarity assessment.

[0060] Word segmentation and encoding: This involves breaking down the sentences in the selection questions into words and encoding each word; word segmentation is a fundamental step in natural language processing. By breaking sentences down into words, we can better understand the structure and semantics of the sentences. Encoding then converts the words into a form that computers can process, facilitating subsequent feature extraction and analysis.

[0061] Generating sentence fingerprints: BERT processes the encoded text to obtain feature vectors for all words, and then averages these word vectors to obtain the sentence fingerprint for selecting questions. BERT is a pre-trained language model capable of understanding the semantics of natural language. By processing the encoded text with BERT, the semantic feature vector of each word can be extracted. Averaging all word vectors yields the semantic feature representation of the entire sentence (i.e., the sentence fingerprint), which provides the foundation for subsequent semantic similarity calculations.

[0062] Similarity Comparison: Calculate the cosine similarity between the sentence fingerprints of two selection questions. Cosine similarity is a commonly used similarity metric, used to calculate the cosine of the angle between two vectors. By calculating the cosine similarity between the fingerprints of two sentences, their semantic similarity can be quantified. If the similarity exceeds a set threshold, the two questions can be considered semantically repetitive.

[0063] As an optional implementation, the final deduplication process determines whether the semantic similarity of the selected questions exceeds a third preset threshold. This is because, in the pre-processing deduplication stage, to save computing power, the pre-deduplication process only removes duplicate questions of the same type. However, the standard question bank still contains duplicate questions of different types. The final deduplication process and the pre-deduplication process work together to achieve the deduplication effect of different types of duplicate questions with less computing power.

[0064] A second aspect of the present disclosure provides an electronic device, characterized in that,

[0065] Including the processor and memory;

[0066] The memory is used to store programs;

[0067] The processor executes the program to implement the method described above.

[0068] A third aspect of the present disclosure provides a computer-readable storage medium, characterized in that,

[0069] The storage medium stores a program, which is executed by a processor to implement the method described above.

[0070] A fourth aspect of this disclosure provides an intelligent document generation device, which includes:

[0071] The data acquisition module is used to collect raw data of various types of questions and to standardize the collected data to form a standard question bank. The standardization process includes adding information such as applicable textbook, applicable grade, question type, first-level knowledge point tag, second-level knowledge point tag, difficulty coefficient, and score to the raw data.

[0072] The preprocessing module is used to pre-deduplicate the questions in the standard question bank, and then divide the standard question bank into multiple first-level question banks according to the first-level knowledge point tags, and divide the first-level question banks into multiple second-level question banks according to the second-level knowledge point tags.

[0073] The test paper creation rule setting module provides a human-computer interaction interface for users to input test paper creation information, and parses and stores the input information;

[0074] The question filtering and matching module filters out questions that meet the set conditions from the question bank based on the set test paper information.

[0075] The test paper generation and adjustment module is used to reasonably allocate the number of each question type and the score range of a single question according to the total score of the test paper and the question type configuration requirements;

[0076] Used to filter questions with corresponding difficulty levels based on a preset difficulty distribution ratio;

[0077] Used to select at least one question from each of the primary question banks associated with the test paper information, and to prioritize selection from the secondary question banks related to core knowledge points;

[0078] This is used to perform final deduplication on the selected questions after the initial test paper compilation is completed. During the final deduplication process, if it is determined that the selected questions are duplicated, it is first determined whether the duplicate questions belong to the secondary question bank related to the core knowledge points. If it is determined that the duplicate questions belong to the secondary question bank related to the core knowledge points, then the duplicate questions are replaced with another question from the same secondary question bank; if it is determined that the duplicate questions do not belong to the secondary question bank related to the core knowledge points, then the duplicate questions are replaced with another question from the same primary question bank.

[0079] The technical effects of this embodiment are as follows: By collecting and standardizing question data, the standard question bank is divided into a primary question bank and a secondary question bank. The primary question bank has a broad knowledge base, while the secondary question bank has precise knowledge points. In the process of test paper assembly, questions are first selected from the primary question bank associated with the test paper information to ensure the comprehensiveness of the test points. Then, the questions in the test paper are finally deduplicated. In the final deduplication process, it is prioritized to determine whether the duplicate questions belong to the secondary question bank related to the core knowledge points. When the duplicate questions belong to the secondary question bank related to the core knowledge points, the duplicate questions are replaced with another question from the same secondary question bank, thereby ensuring the accuracy of test paper assembly again.

[0080] The above are merely preferred embodiments of this disclosure and are not intended to limit this disclosure. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. An intelligent test paper compiling method, characterized in that, The intelligent test paper compiling method comprises: Collecting original data of various types of questions and performing standardization processing on the collected data to form a standard question bank; wherein the standardization processing comprises adding applicable teaching materials, applicable grades, question types, first knowledge point labels, second knowledge point labels, difficulty coefficients and score information to the original data; Performing pre-deduplication processing on the questions in the standard question bank, and then dividing the standard question bank into a plurality of first question banks according to the first knowledge point labels, and dividing the first question banks into a plurality of second question banks according to the second knowledge point labels; According to the set test paper compiling information, questions are screened from the pre-deduplicated standard question bank, and the specific process is as follows: According to the total score of the test paper and the question type configuration requirements, the number of each type of question and the single question score interval are reasonably allocated; According to the pre-set difficulty distribution ratio, questions with corresponding difficulty coefficients are screened; At least one question is screened from each first question bank associated with the test paper compiling information, and the screening is preferentially performed from the second question bank related to the core knowledge point; After the test paper compilation is initially completed, final deduplication processing is performed on the screened questions. In the final deduplication processing, if it is judged that the screened questions are duplicated, it is preferentially judged whether the duplicated questions belong to the second question bank related to the core knowledge point. If it is judged that the duplicated questions belong to the second question bank related to the core knowledge point, the duplicated questions are replaced by another question in the same second question bank. If it is judged that the duplicated questions do not belong to the second question bank related to the core knowledge point, the duplicated questions are replaced by another question in the same first question bank; The final deduplication processing comprises a first detection means for detecting whether the question type and the second knowledge point label of the screened questions are duplicated. When the question type and the second knowledge point label of two questions in the screened questions are completely the same, the first detection means judges that the screened questions are duplicated in knowledge points; The final deduplication processing further comprises a second detection means for judging whether the similarity of the answers and the problem solving steps of the screened questions exceeds a first set threshold value. When the similarity of the answers and the problem solving steps of two questions in the screened questions exceeds the first set threshold value, the second detection means judges that the screened questions are duplicated in knowledge points; wherein, when the first detection means judges that the screened questions are duplicated in knowledge points, the second detection means is entered; when the first detection means judges that the screened questions are not duplicated in knowledge points, the second detection means is not entered; The final deduplication processing further comprises a third detection means for judging whether the similarity of the semantics of the screened questions exceeds a second set threshold value. When the similarity of the semantics of two questions in the screened questions exceeds the second set threshold value, the third detection means judges that the screened questions are duplicated in knowledge points; wherein, when the second detection means judges that the screened questions are duplicated in knowledge points, the third detection means is entered; when the second detection means judges that the screened questions are not duplicated in knowledge points, the third detection means is not entered.

2. The intelligent test assembly method of claim 1, wherein, The primary knowledge point label includes a plurality of associated secondary knowledge point labels.

3. The intelligent test assembly method of claim 1, wherein, The specific steps of judging the similarity of the semantic of the screening questions by the third detection means include: Cleaning the screening questions: removing irrelevant content in the screening questions and unifying the format; Word segmentation and encoding: splitting the sentences in the screening questions into words and encoding each word; Generating sentence fingerprints: processing the encoded text by BERT to obtain the feature vectors of all words, and then averaging all word vectors to obtain the sentence fingerprint of the screening question; Comparing the similarity: calculating the cosine similarity between the sentence fingerprints of two screening questions.

4. The intelligent test assembly method of claim 3, wherein, The final deduplication processing judges whether the similarity of the semantic of the screening questions exceeds a third set threshold.

5. An electronic device, characterized in that, comprising a processor and a memory; the memory is used to store a program; the processor executes the program to realize the method of any one of claims 1-4.

6. A computer readable storage medium, characterized in that, the storage medium stores a program, and the program is executed by a processor to realize the method of any one of claims 1-4.

7. An intelligent test paper compiling device, characterized in that, The intelligent test paper setting device comprises: a data acquisition module, configured to acquire original data of various types of questions and perform standardization processing on the acquired data to form a standard question bank; wherein the standardization processing comprises adding applicable teaching materials, applicable grades, question types, primary knowledge point labels, secondary knowledge point labels, difficulty coefficients and score information to the original data; a preprocessing module, configured to perform pre-deduplication processing on the questions in the standard question bank, and then divide the standard question bank into a plurality of primary question banks according to the primary knowledge point labels, and divide the primary question banks into a plurality of secondary question banks according to the secondary knowledge point labels; a test paper setting rule setting module, configured to provide a man-machine interactive interface for a user to input test paper setting information, and analyze and store the input information; a question screening and matching module, configured to screen questions meeting the conditions from the standard question bank after the pre-deduplication processing according to the set test paper setting information; a test paper generation and adjustment module, configured to reasonably allocate the number of each type of question and the score interval of each question according to the total score of the test paper and the question type configuration requirements; configured to screen questions corresponding to the difficulty coefficients according to the preset difficulty distribution ratio; configured to screen at least one question from each of the primary question banks associated with the test paper setting information, and preferentially screen from the secondary question banks related to the core knowledge points; configured to perform final deduplication processing on the screened questions after the initial test paper setting is completed, and preferentially judge whether the repeated questions belong to the secondary question banks related to the core knowledge points in the final deduplication processing, replace the repeated questions with another question in the same secondary question bank if the repeated questions belong to the secondary question banks related to the core knowledge points, and replace the repeated questions with another question in the same primary question bank if the repeated questions do not belong to the secondary question banks related to the core knowledge points. The final deduplication processing includes a first detection means for detecting whether the question type and the secondary knowledge point label of the screening questions are duplicated, and determining that the screening questions have duplicated knowledge points when the question type and the secondary knowledge point label of two questions in the screening questions are completely identical; The final deduplication processing also includes a second detection means for determining whether the similarity of the answers and the problem solving steps of the screening questions exceeds a first set threshold, and determining that the screening questions have duplicated knowledge points when the similarity of the answers and the problem solving steps of two questions in the screening questions exceeds the first set threshold; wherein, when the first detection means determines that the screening questions have duplicated knowledge points, the second detection means is entered; and when the first detection means determines that the screening questions do not have duplicated knowledge points, the second detection means is not entered; The final deduplication processing also includes a third detection means for determining whether the similarity of the semantics of the screening questions exceeds a second set threshold, and determining that the screening questions have duplicated knowledge points when the similarity of the semantics of two questions in the screening questions exceeds the second set threshold; wherein, when the second detection means determines that the screening questions have duplicated knowledge points, the third detection means is entered; and when the second detection means determines that the screening questions do not have duplicated knowledge points, the third detection means is not entered.

Images