Accurate test question duplicate checking method and system

Abstract

The invention relates to the technical field of test question duplicate checking, and particularly discloses an accurate test question duplicate checking method and system.The method comprises the steps that a test question database is constructed, and question information and answer information of test questions are recorded and stored; processing the test question information to obtain duplicate checking test questions, preprocessing the duplicate checking test questions, and then performing statement decomposition and extraction on the text information by taking punctuation marks as intervals; performing traversal duplicate checking comparison on the duplicate checking test questions and questions in the test question database to obtain duplicate signals; the method comprises the following steps: marking and displaying target questions in a database with repeated signals obtained through traversal duplicate checking, firstly comparing the number of words of duplicate checking test questions with the number of words of the target questions in the database so as to preliminarily judge the overlap ratio of the test questions, and when the overlap ratio of keywords is relatively high, indicating that the meanings of the duplicate checking test questions are similar to those of the target test questions, and otherwise, indicating that the meanings of the duplicate checking test questions are similar to those of the target test questions. And calculating and judging whether the duplicate checking test question and the target test question are duplicate based on the test question text statement total value difference and the text statement meaning similarity ratio of the duplicate checking test question.

Description

Technical Field

[0001] The present invention relates to the technical field of test question duplication checking, and in particular to a method and system for accurately checking test question duplication. Background Art

[0002] There are usually two existing methods for checking duplicate questions: one is to use the inverted index to check duplicate questions. Although this method has a faster search speed, it has the following three disadvantages: first, it does not support checking duplicate questions when some documents are updated; second, if new documents are to be searched, the entire inverted index needs to be updated, which limits the amount of data contained in the index and the frequency of index updates; third, it does not support checking duplicate questions based on the semantics of the test questions. Another commonly used method for checking duplicate questions is to use algorithms to check duplicate questions;

[0003] The above-mentioned duplicate checking method can only check for duplicates based on the same text during the duplicate checking process, and the duplicate checking effect is not good: it leads to low accuracy of the duplicate checking results and the entire duplicate checking speed is too slow. Summary of the invention

[0004] The purpose of the present invention is to provide a method and system for accurately checking for duplicate test questions to solve the above-mentioned problems.

[0005] The purpose of the present invention can be achieved through the following technical solutions:

[0006] A method for accurately checking for duplicate test questions, comprising the following steps:

[0007] Step 1: Build a test database to collect and store test question information and answer information;

[0008] The test information is processed as follows: the test information is stripped of text to obtain text information; the text information is decomposed and extracted into sentences with punctuation marks as intervals; and keywords are extracted from each decomposed sentence;

[0009] Step 2: Obtain duplicate test questions and pre-process the duplicate test questions. The processing process includes: cleaning the test question information to obtain text information, and then decomposing and extracting sentences based on punctuation marks;

[0010] Step 3: traverse and compare the duplicate test questions with the test question database to obtain duplicate signals;

[0011] Step 4: Mark and display the target questions in the database that have obtained duplicate signals through traversal checking.

[0012] As a further solution of the present invention: in step 1, the text information includes: the question text, that is, the text information of each test question in the database, including: the total value KZ of the question text sentence; and the question text sentence word value KS in each text sentence i , and get the keyword quantity value of each question text sentence and mark it as KG i ; Wherein, i is 1, 2, 3…, and i is a positive integer; i is used for sentence markers in the question text.

[0013] As a further solution of the present invention: in the step 2, the text information is decomposed and extracted: including obtaining the total value of the text sentence in the duplicate check text and the value of each sentence word in the duplicate check test text; and marking them as CCZ and CCS respectively j ; Wherein, j is 1, 2, 3..., and j is a positive integer; j is used for the sentence marker in the text of the duplicate check question;

[0014] At the same time, keyword extraction is performed on the repeated test questions, and the keyword information of each question text sentence in the text information is extracted, including keywords and the number of keywords, and marked as CCG j .

[0015] As a further solution of the present invention: in the step 3, the duplicate test questions are traversed and compared with the questions in the test question database, including:

[0016] A1: Obtain the data information of the target question from the database, calculate the total difference value ZZC of the test question text, and obtain the first-level duplicate checking signal;

[0017] A2: Based on the primary duplicate check signal, compare the keywords in each sentence of the duplicate check test with the keywords in the target question text, and obtain the number of keyword matches in each question text sentence, and mark it as PP i ;

[0018] Calculate the similarity ratio of the meaning of the text sentences in each question text in the repeated test questions WBS i ;

[0019] Then calculate and obtain the mean WSBJ of the similarity ratio of the meaning of the text sentences;

[0020] Compare the mean value WSBJ of the similarity ratio of the meaning of the text sentences with the threshold value WSBY of the similarity ratio of the meaning of the text sentences; obtain the secondary duplicate checking signal;

[0021] A3: Based on the secondary duplicate checking signal: Get the word value CCS of the text sentence in the duplicate checking test question j and the word value KS of the target question text sentence i ; Calculate the comprehensive ratio of duplicate checking CZHB, and then get the duplicate signal.

[0022] As a further solution of the present invention: in said A1, the total difference value ZZC of the test question text sentence is compared with the total difference threshold value ZZCy of the test question text sentence;

[0023] If 0≤ total difference value ZZC of test text sentences≤ total difference threshold value ZZCy of test text sentences; generate a first-level duplicate checking signal;

[0024] If the total value difference ZZC of the test text statement is greater than the total value difference threshold ZZCy of the test text statement, a difference signal is generated.

[0025] As a further solution of the present invention: in A2, if the text sentence meaning similarity ratio mean WSBJ is greater than or equal to the text sentence meaning similarity ratio threshold WSBY, a secondary duplicate checking signal is generated;

[0026] If the text sentence meaning similarity ratio mean WSBJ is less than the text sentence meaning similarity ratio threshold WSBY, a difference signal is generated.

[0027] As a further solution of the present invention: the calculation process of the comprehensive ratio of duplicate checking CZHB includes:

[0028] Obtain the meaning deviation value HP of each text sentence in the repetition test question by calculation j ;

[0029] Then calculate and obtain the total meaning deviation HPZ of the text sentences in the recheck test questions;

[0030] The comprehensive ratio of duplicate checking CZHB is obtained by calculation, where a is the sentence difference coefficient, b is the meaning similarity coefficient, and c is the meaning deviation coefficient.

[0031] As a further solution of the present invention: comparing the comprehensive ratio of duplicate checking CZHB with the comprehensive ratio threshold of duplicate checking CZHBy;

[0032] If the comprehensive ratio of duplicate checking CZHB is greater than or equal to the comprehensive ratio threshold of duplicate checking CZHBy, a duplicate signal is generated;

[0033] If the comprehensive ratio of plagiarism checking CZHB is less than the comprehensive ratio threshold of plagiarism checking CZHBy, a difference signal is generated.

[0034] As a further solution of the present invention: in the step 4, based on the repetitive signal, the answer information of the target test question can also be displayed.

[0035] As a further solution of the present invention: a system for accurately checking duplicate test questions, comprising:

[0036] Database construction module: used to construct the test question database, collect and store the question information and answer information of the test questions;

[0037] The test information is processed as follows: the test information is stripped of text to obtain text information; the text information is decomposed and extracted into sentences with punctuation marks as intervals; and keywords are extracted from each decomposed sentence;

[0038] Data acquisition module: used to obtain duplicate test questions and pre-process the duplicate test questions. The processing process includes: cleaning the test question information to obtain text information, and then decomposing and extracting sentences based on punctuation marks;

[0039] Duplicate check module: traverse and compare the duplicate test questions with the test questions database to obtain duplicate signals;

[0040] Display module: used to mark and display the target questions in the database that have obtained duplicate signals through traversal checking.

[0041] Beneficial effects of the present invention:

[0042] In the present invention, the word count of the duplicate test question is first compared with the word count of the target question in the database to make a preliminary judgment on the overlap of the test questions, and then the keywords in each sentence of the duplicate test question are compared with the keywords in the target test question; when the keyword overlap is high, it means that the meaning of the duplicate test question and the target test question are similar, based on this, the sentences of the duplicate test question and the target test question are compared, and further based on the total value difference of the test text sentences of the duplicate test question and the similarity ratio of the text sentence meaning, it is calculated and judged whether the duplicate test question and the target test question are repeated, and then it is judged whether the test question database contains questions similar to the duplicate test question; the accuracy is higher and the effect is better;

[0043] The duplicate checking of test questions in the database can be used by teachers to update the test question bank, and can be used to retrieve test questions during the students' learning process to ensure that the students' learning process can be better consolidated based on the same type of test questions; it can also be used to search for test answer analysis, etc. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] The present invention will be further described below in conjunction with the accompanying drawings.

[0045] Figure 1 It is a schematic diagram of the process structure of the method of the present invention;

[0046] Figure 2 It is a schematic diagram of the process flow of the duplicate checking method in the present invention;

[0047] Figure 3 It is a system block diagram of the present invention. DETAILED DESCRIPTION

[0048] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0049] Embodiment 1

[0050] See also Figure 1 As shown, the present invention is a method for accurately checking duplicate test questions, comprising the following steps:

[0051] Step 1: Build a test database to collect and store test question information and answer information;

[0052] The test information is processed as follows: the test information is stripped of text to obtain text information; the text information is decomposed and extracted using punctuation marks as intervals; and keywords are extracted from each decomposed sentence. It should be noted that in the process of extracting keywords from text information, the Word2Vec technology is used to extract keywords from each sentence.

[0053] The text information includes the question text, that is, the text information of each test question in the database, including: the total value of the question text sentence KZ; and the question text sentence word value KS in each text sentence i , and get the keyword quantity value of each question text sentence and mark it as KG i ; Wherein, i is 1, 2, 3, ..., and i is a positive integer; i is used for sentence markers in the question text;

[0054] Step 2: Get the duplicate test questions and pre-process them. The processing process includes: cleaning the test question information to obtain text information, then decomposing and extracting sentences based on punctuation marks; obtaining the total sentence value of the duplicate text and the word value of each sentence in the duplicate test question text; and marking them as CCZ and CCS respectively. j ; Wherein, j is 1, 2, 3..., and j is a positive integer; j is used for the sentence marker in the text of the duplicate check question;

[0055] At the same time, keyword extraction is performed on the repeated test questions, and the keyword information of each question text sentence in the text information is extracted, including keywords and the number of keywords, and marked as CCG j ;

[0056] Step 3: traverse and compare the duplicate test questions with the test database questions; Figure 2 As shown, specifically including:

[0057] A1: Get the target topic data from the database. Calculate and obtain the total difference value ZZC of the test text sentences;

[0058] Compare the total difference value ZZC of the test text sentences with the total difference threshold value ZZCy of the test text sentences;

[0059] If 0≤ total difference value ZZC of test text sentences≤ total difference threshold value ZZCy of test text sentences; generate a first-level duplicate checking signal;

[0060] If the total difference value ZZC of the test text sentence is greater than the total difference threshold value ZZCy of the test text sentence, a difference signal is generated;

[0061] A2: Based on the primary duplicate check signal, compare the keywords in each sentence of the duplicate check test with the keywords in the target question text, and obtain the number of keyword matches in each question text sentence, and mark it as PP i ;

[0062] pass Calculate the similarity ratio of the meaning of the text sentences in each question text in the repeated test questions WBS i ;

[0063] Then through Calculate and obtain the mean WSBJ of the similarity ratio of the meaning of the text sentences;

[0064] Compare the mean value of the similarity ratio of text sentences WSBJ with the threshold value of the similarity ratio of text sentences WSBY;

[0065] If the mean value of the similarity ratio of the text sentence meaning WSBJ is greater than or equal to the threshold value of the similarity ratio of the text sentence meaning WSBY, a secondary duplicate checking signal is generated;

[0066] If the mean value WSBJ of the similarity ratio of the meaning of the text sentences is less than the threshold value WSBY of the similarity ratio of the meaning of the text sentences, a difference signal is generated;

[0067] A3: Based on the secondary duplicate checking signal: Get the word value CCS of the text sentence in the duplicate checking test question j and the word value KS of the target question text sentence i ;

[0068] pass Calculate the meaning deviation value HP of each text sentence in the repetition test question j ;

[0069] Then through HPZ = ∑HP j Calculate and obtain the total meaning deviation HPZ of the text sentences in the re-examination questions;

[0070] pass The comprehensive ratio of duplicate checking CZHB is calculated, where a is the sentence difference coefficient, b is the meaning similarity coefficient, and c is the meaning deviation coefficient;

[0071] Compare the comprehensive ratio of duplicate checking CZHB with the comprehensive ratio threshold of duplicate checking CZHBy;

[0072] If the comprehensive ratio of duplicate checking CZHB is greater than or equal to the comprehensive ratio threshold of duplicate checking CZHBy, a duplicate signal is generated;

[0073] If the comprehensive ratio of duplicate checking CZHB is less than the comprehensive ratio threshold CZHBy, a difference signal is generated;

[0074] It should be noted that when traversing the test database for duplicate checking, after obtaining the test questions with duplicate signals, it is still necessary to conduct subsequent duplicate checking on the remaining test questions in the test database, so as to display all the test questions in the test database that are similar to the duplicate checking test questions for the duplicate checking personnel to view;

[0075] Step 4: Mark and display the target questions in the database that have obtained duplicate signals after traversing the duplicate check; at the same time, the answer information of the target test question can also be displayed;

[0076] First, the word count of the duplicate test question is compared with the word count of the target question in the database to make a preliminary judgment on the overlap of the test questions. Then, the keywords in each sentence of the duplicate test question are compared with the keywords in the target test question. When the keyword overlap is high, it means that the meaning of the duplicate test question and the target test question are similar. Based on this, the sentences of the duplicate test question and the target test question are compared. Based on the total value difference of the test text sentences of the duplicate test question and the similarity ratio of the text sentence meaning, it is calculated and judged whether the duplicate test question and the target test question are repeated, and then it is judged whether the test question database contains questions similar to the duplicate test question. It has higher accuracy and better effect.

[0077] The duplicate checking of test questions in the database can be used by teachers to update the test question bank, and can be used to retrieve test questions during the students' learning process to ensure that the students' learning process can be better consolidated based on the same type of test questions; it can also be used to search for test answer analysis, etc.

[0078] Embodiment 2

[0079] like Figure 3 As shown, a system for accurately checking duplicate test questions includes:

[0080] Database construction module: used to construct the test question database, collect and store the question information and answer information of the test questions;

[0081] The test information is processed as follows: the test information is stripped of text to obtain text information; the text information is decomposed and extracted into sentences with punctuation marks as intervals; and keywords are extracted from each decomposed sentence;

[0082] Data acquisition module: used to obtain duplicate test questions and pre-process the duplicate test questions. The processing process includes: cleaning the test question information to obtain text information, and then decomposing and extracting sentences based on punctuation marks;

[0083] Duplicate check module: traverse and compare the duplicate test questions with the test questions database to obtain duplicate signals;

[0084] Display module: used to mark and display the target questions in the database that have obtained duplicate signals through traversal checking.

[0085] The word count of the repetition test question is compared with the word count of the target question in the database to make a preliminary judgment on the overlap of the test questions. Then, the keywords in each sentence of the repetition test question are compared with the keywords in the target test question. When the keyword overlap is high, it means that the meaning of the repetition test question and the target test question are similar. Based on this, the sentences of the repetition test question and the target test question are compared. Based on the total value difference of the test text sentences of the repetition test question and the similarity ratio of the text sentence meaning, it is further calculated and judged whether the repetition test question and the target test question are repeated, and then it is judged whether the test question database contains questions similar to the repetition test question.

[0086] The duplicate checking of test questions in the database can be used by teachers to update the test question bank, and can be used to retrieve test questions during the students' learning process to ensure that the students' learning process can be better consolidated based on the same type of test questions; it can also be used to search for test answer analysis, etc.

[0087] In the description of this specification, the description with reference to the terms "one embodiment", "example", "specific example", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0088] The above formulas are all dimensionless and numerical calculations. The formula is a formula obtained by collecting a large amount of data and performing software simulation to obtain the most recent real situation. The preset parameters in the formula are set by technicians in this field according to actual conditions. The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to only specific implementation methods. Obviously, many modifications and changes can be made according to the contents of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that technicians in the relevant technical field can understand and use the present invention well. The present invention is only limited by the claims and their full scope and equivalents.

[0089] The above is a detailed description of an embodiment of the present invention, but the content is only a preferred embodiment of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. A method for accurately checking for duplicate test questions, characterized in that: The following steps are involved: Step 1: Build a test database to collect and store test question information and answer information; The test information is processed as follows: the test information is stripped of text to obtain text information; the text information is decomposed and extracted into sentences with punctuation marks as intervals; and keywords are extracted from each decomposed sentence; Step 2: Obtain duplicate test questions and pre-process the duplicate test questions. The processing process includes: cleaning the test question information to obtain text information, and then decomposing and extracting sentences based on punctuation marks; Step 3: traverse and compare the duplicate test questions with the test question database to obtain duplicate signals; Step 4: Mark and display the target questions in the database that have obtained duplicate signals through traversal checking.

2. According to claim 1, a method for accurately checking for duplicate examination questions is characterized in that: In step 1, the text information includes: the question text, that is, the text information of each test question in the database, including: the total value KZ of the question text sentence; and the question text sentence word value KS in each text sentence i , and get the keyword quantity value of each question text sentence and mark it as KG i ; Wherein, i is 1, 2, 3…, and i is a positive integer; i is used for sentence markers in the question text.

3. A method for accurately checking for duplicate examination questions according to claim 2, characterized in that: In the step 2, the text information is decomposed and extracted: including obtaining the total value of the text sentence and the word value of each sentence in the duplicate text; and marking them as CCZ and CCS respectively. j ; Wherein, j is 1, 2, 3..., and j is a positive integer; j is used for the sentence marker in the text of the duplicate check question; At the same time, keyword extraction is performed on the repeated test questions, and the keyword information of each question text sentence in the text information is extracted, including keywords and the number of keywords, and marked as CCG j .

4. A method for accurately checking for duplicate examination questions according to claim 3, characterized in that: In the step 3, the duplicate test questions are traversed and compared with the test questions in the test question database, including: A1: Obtain the data information of the target question from the database, calculate the total difference value ZZC of the test question text sentences, and obtain the first-level duplicate checking signal; A2: Based on the primary duplicate check signal, compare the keywords in each sentence of the duplicate check test with the keywords in the target question text, and obtain the number of keyword matches in each question text sentence, and mark it as PP i ; Calculate the similarity ratio of the meaning of the text sentences in each question text in the repeated test questions WBS i ; Then calculate and obtain the mean WSBJ of the similarity ratio of the meaning of the text sentences; Compare the mean value WSBJ of the similarity ratio of the meaning of the text sentences with the threshold value WSBY of the similarity ratio of the meaning of the text sentences; obtain the secondary duplicate checking signal; A3: Based on the secondary duplicate check signal: Get the CCS value of the text sentence in the duplicate check question j and the word value KS of the target question text sentence i ; Calculate the comprehensive ratio of duplicate checking CZHB, and then get the duplicate signal.

5. A method for accurately checking for duplicate examination questions according to claim 4, characterized in that: In A1, the total difference value ZZC of the test text sentence is compared with the total difference threshold value ZZCy of the test text sentence; If 0≤ total difference value ZZC of test text sentences≤ total difference threshold value ZZCy of test text sentences; Generate a first-level duplicate checking signal; If the total value difference ZZC of the test text statement is greater than the total value difference threshold ZZCy of the test text statement, a difference signal is generated.

6. A method for accurately checking for duplicate examination questions according to claim 4, characterized in that: In A2, if the text sentence meaning similarity ratio mean WSBJ is greater than or equal to the text sentence meaning similarity ratio threshold WSBY, a secondary duplicate checking signal is generated; If the text sentence meaning similarity ratio mean WSBJ is less than the text sentence meaning similarity ratio threshold WSBY, a difference signal is generated.

7. A method for accurately checking for duplicate examination questions according to claim 4, characterized in that: The calculation process of the comprehensive ratio of duplicate checking CZHB includes: Calculate the meaning deviation value HP of each text sentence in the repetition test question j ; Then through HPZ = ∑HP j Calculate and obtain the total meaning deviation HPZ of the text sentences in the re-examination questions; The comprehensive ratio of duplicate checking CZHB is calculated, where a is the sentence difference coefficient, b is the meaning similarity coefficient, and c is the meaning deviation coefficient.

8. A method for accurately checking for duplicate examination questions according to claim 7, characterized in that: Compare the comprehensive ratio of duplicate checking CZHB with the comprehensive ratio threshold of duplicate checking CZHBy; If the comprehensive ratio of duplicate checking CZHB is greater than or equal to the comprehensive ratio threshold of duplicate checking CZHBy, a duplicate signal is generated; If the comprehensive ratio of plagiarism checking CZHB is less than the comprehensive ratio threshold of plagiarism checking CZHBy, a difference signal is generated.

9. A method for accurately checking for duplicate examination questions according to claim 1, characterized in that: In the step 4, based on the repetition signal, the answer information of the target test question can also be displayed.

10. A system for accurately checking for duplicate examination questions, characterized in that: The system is used to execute the method for accurately checking for duplicate examination questions as described in any one of claims 1 to 9, comprising: Database construction module: used to construct the test question database, collect and store the question information and answer information of the test questions; The test information is processed as follows: the test information is stripped of text to obtain text information; the text information is decomposed and extracted into sentences with punctuation marks as intervals; and keywords are extracted from each decomposed sentence; Data acquisition module: used to obtain duplicate test questions and pre-process the duplicate test questions. The processing process includes: cleaning the test question information to obtain text information, and then decomposing and extracting sentences based on punctuation marks; Duplicate check module: traverse and compare the duplicate test questions with the test questions database to obtain duplicate signals; Display module: used to mark and display the target questions in the database that have obtained duplicate signals through traversal checking.

Images