Chinese examination system suitable for learners with diverse backgrounds, and settings of actual operation and feedback thereof
The Chinese examination system dynamically adjusts question difficulty and provides real-time feedback to learners with diverse backgrounds, addressing the limitations of existing systems by offering automated and equitable assessment.
Patent Information
- Application Number
- GB2026008976
- Authority / Receiving Office
- GB · GB
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2023-10-17
- Filing Date
- 2024-09-24
- Publication Date
- 2026-11-11
AI Technical Summary
Existing Chinese examination systems fail to accommodate learners with diverse backgrounds by either being targeted for either non-native or native speakers or native speakers, lack intelligent recognition, and provide delayed feedback with limited presentation forms.
A Chinese examination system and method that includes a server, client, and data transmission network, utilizing an adaptive module question bank and diagnostic module question bank to dynamically adjust question difficulty based on examinee responses, providing real-time feedback and personalized assessment.
The system offers automated, efficient, and equitable assessment of Chinese proficiency across diverse backgrounds, reducing human intervention, enhancing accuracy, and providing personalized and timely feedback.
Smart Images

Figure 00000000_0000_ABST
Abstract
Description
[0001] The present application relates to the technical field of examination systems, and in particular, to a Chinese examination system suitable for learners with diverse backgrounds, and settings of actual operation and feedback thereof. More specifically, the present application relates to a Chinese examination system and method suitable for learners with diverse backgrounds. BACKGROUND
[0002] Currently, Chinese examination systems are either designed for non-native Chinese speakers or targeted at native Chinese speakers, which cannot achieve both simultaneously or perform intelligent recognition and push. Moreover, examinations conducted in the same exam mostly employ a unified question bank, present delayed exam feedback and still require manual assistance, and the presentation form of exam results is relatively single.
[0003] The patent document CN112581089A (Application No.: 202011508998.8) discloses a system based on the civil aviation certification examination process, comprising: a login module for verifying examinee information; a subject management module that automatically matches examination subjects based on the type of certificate to be sought and establishes a test question bank; a question management module that generates a test paper through test paper composition management; an examination planning module that develops an examination plan based on examinee information such as number and region; an examination scheduling management module that generates proctor information based on a preset matching principle and imports invigilator information; an examination module that stores examination answers and limits submission time; a proctoring management module that monitors the examination module in real time; a score management module that calculates scores according to a preset scoring rule; and a system management module that utilizes an organizational management function to manage the hierarchical relationships of each role, thereby achieving permission isolation. SUMMARY
[0004] With regard to disadvantages of the prior art, an object of the present disclosure is to provide a Chinese examination system suitable for learners with diverse backgrounds, and settings of actual operation and feedback thereof. More specifically, the present disclosure relates to a Chinese examination system and method suitable for learners with diverse backgrounds.
[0005] A Chinese examination system suitable for learners with diverse backgrounds according to the present disclosure comprises: a server, a client, and a data transmission network;
[0006] wherein the client is a terminal device for a Chinese examination;
[0007] the server provides a database service for the Chinese examination and comprises a test question information database, the test question information database comprising an adaptive module question bank and a diagnostic module question bank;
[0008] the data transmission network is connected to the server and the client and is used for guaranteeing data transmission between the server and the client; and
[0009] when the Chinese examination begins, the server pushes the adaptive module question bank of a corresponding level on the basis of examinee information provided by the client; the client transmits corresponding responses to the server, the server determines the difficulty level of the next round of test questions on the basis of the responses provided by the client and pushes the next round of test questions to the client, until the error rate of the responses of the adaptive module question bank of the corresponding level meets a preset condition, and obtains the current examinee’s Chinese proficiency level; the server pushes the diagnostic module question bank corresponding to the current examinee’s level to the client, the client feeds back the corresponding responses to the server, and the server performs a diagnosis assessment of the current examinee’s proficiency level on the basis of the responses.
[0010] Preferably, the server pushes a test question version on the basis of a font for examination selected by the client.
[0011] Preferably, the server pushes the adaptive module question bank of the corresponding level on the basis of the examinee information provided by the client, the server determines the responses provided by the client, pushes the adaptive module question bank of a higher level to the client when the accuracy rate meets a preset requirement, and obtains the current examinee’s Chinese proficiency level when the accuracy rate does not meet the preset requirement, and continues to push the diagnostic module test question bank of the current Chinese proficiency level to the client.
[0012] Preferably, the server counts scores and erroneous options from examinees’ responses, performs an error attribution, and generates a corresponding report for the Chinese proficiency level, real-time ranking, and ability description.
[0013] Preferably, scores and erroneous options from the responses of all examinees are counted and score reports organized in units of class or the whole school are automatically generated.
[0014] A Chinese examination method suitable for learners with diverse backgrounds according to the present disclosure comprises:
[0015] Step SI: where a server pushes an adaptive module question bank of a corresponding level on the basis of examinee information provided by a client;
[0016] Step S2: where the client transmits corresponding responses to the server, the server determines the difficulty level of the next round of test questions on the basis of the responses provided by the client and pushes the next round of test questions to the client, until the error rate of the responses of the adaptive module question bank of the corresponding level meets a preset condition, and obtains the current examinee’s Chinese proficiency level;
[0017] Step S3: where the server pushes a diagnostic module question bank corresponding to the current examinee’s level to the client, the client feeds back the corresponding responses to the server, and the server performs a diagnosis assessment of the current examinee’s proficiency level on the basis of the responses.
[0018] Preferably, the server pushes a test question version on the basis of a font for examination selected by the client.
[0019] Preferably, the server pushes the adaptive module question bank of the corresponding level on the basis of the examinee information provided by the client, the server determines the responses provided by the client, pushes the adaptive module question bank of a higher level to the client when the accuracy rate meets a preset requirement, and obtains the current examinee’s Chinese proficiency level when the accuracy rate does not meet the preset requirement, and continues to push the diagnostic module test question bank of the current Chinese proficiency level to the client.
[0020] Preferably, the server counts scores and erroneous options from examinees’ responses, automatically performs an error attribution, and generates a corresponding report for the Chinese proficiency level, real-time ranking, and ability description.
[0021] Preferably, scores and erroneous options from examinees’ responses of all examinees are counted and score reports organized in units of class or the whole school are automatically generated.
[0022] As compared with the prior art, the present disclosure provides the following beneficial effects:
[0023] 1. The Chinese examination system and method suitable for learners with diverse backgrounds according to the present disclosure can be fully automated, requiring no manual intervention or judgement. This reduces human resource costs, simplifies test administration, and enables a comprehensive and efficient assessment of the ability to grasp Chinese skills;
[0024] 2. Enhanced Testing Accuracy: Traditional fixed-length tests can only provide a rough assessment of the test-taker’s ability on the basis of the difficulty of questions, whereas the present disclosure dynamically selects test questions of appropriate difficulty on the basis of the test-taker’s responses, thereby more precisely evaluating the test-taker’s proficiency level;
[0025] 3. Reduced Testing Time: Traditional fixed-length tests typically require longer testing time, whereas the present disclosure automatically adjusts the test length on the basis of the test-taker’s proficiency level and testing objectives, thereby saving the testing time and improving the testing efficiency;
[0026] 4. Enhanced Test Applicability: Due to regional and cultural variations in Chinese languages, traditional fixed-length tests may exhibit linguistic and cultural biases, whereas the present disclosure dynamically selects appropriate test questions on the basis of background information of diverse learners and test-takers and the testing objectives, thereby enhancing the test applicability;
[0027] 5. Personalized Assessment: The present disclosure can provide personalized tests for test-takers of varying proficiency levels on the basis of their performance, thereby more accurately assessing their proficiency levels across different dimensions and effectively measuring the language proficiency of Chinese learners across different courses and age groups;
[0028] 6. By adopting a unified, standardized, and intelligent Chinese examination system, the present disclosure not only frees teachers from the tasks of creating, administering, and grading exams, but also ensures that testing for learners with diverse backgrounds is more equitable and standardized, thus providing a powerful tool for promoting Chinese language education both domestically and internationally;
[0029] 7. By adopting three interactive modes—school, teacher, and individual student—to present a rich score report chart on campuses, classes, and individuals, the present disclosure addresses the problem of limited application of score data in teaching, which arises from most school administrators and teachers lacking professional data analysis skills. BRIEF DESCRIPTION OF THE DRAWINGS
[0030] Other features, objects, and advantages of the present disclosure will become more apparent upon reading the detailed description of non-limiting embodiments with reference to the accompanying drawings:
[0031] FIG. lisa flowchart of a Chinese examination method suitable for learners with diverse backgrounds. DETAILED DESCRIPTION
[0032] The present disclosure will be described in detail below in conjunction with specific embodiments. The following embodiments will assist those skilled in the art to further understand the present disclosure, but do not limit the scope of the present disclosure in any way. It should be noted that, for one of ordinary skill in the art, several modifications and improvements could be made without departing from the spirit of the present disclosure. These are all within the scope of protection of the present disclosure.
[0033] Learners with diverse backgrounds take exams on the same testing platform using the same evaluation criteria. Examinees’ responses are automatically pushed to a report platform database, and the report platform updates the examinees’ proficiency levels, ability diagnostic results, and dynamic rankings in real time. Unified, standardized, and intelligent vocabulary examinations not only free teachers from the tasks of creating, administering, and grading exams as well as analyzing scores, but also ensure that testing for learners with diverse backgrounds is more equitable and standardized, thus providing a powerful tool for promoting Chinese language education both domestically and internationally.
[0034] Embodiment 1
[0035] A Chinese examination system suitable for learners with diverse backgrounds according to the present disclosure, as shown in FIG. 1, comprises: a server, a client, and a data transmission network;
[0036] wherein the client is a terminal device for a Chinese examination;
[0037] the server provides a database service for the Chinese examination and comprises a test question information database, the test question information database comprising an adaptive module question bank and a diagnostic module question bank;
[0038] the data transmission network is connected to the server and the client and is used for guaranteeing data transmission between the server and the client;
[0039] when the Chinese examination is started, the server pushes the adaptive module question bank of a corresponding level on the basis of examinee information provided by the client; the client transmits the corresponding responses to the server, the server determines the difficulty level of the next round of test questions on the basis of the responses provided by the client and pushes the next round of test questions to the client, until the error rate of the responses of the adaptive module question bank of the corresponding level meets a preset condition, and obtains the current examinee’s Chinese proficiency level; the server pushes the diagnostic module question bank corresponding to the current examinee’s level to the client, the client feeds back the corresponding responses to the server, and the server performs a diagnosis assessment of the current examinee’s proficiency level on the basis of the responses.
[0040] Specifically, the server pushes a test question version on the basis of a font for examination selected by the client.
[0041] Specifically, the server pushes the adaptive module question bank of the corresponding level on the basis of the examinee information provided by the client, the server determines the responses provided by the client, pushes the adaptive module question bank of a higher level to the client when the accuracy rate meets a preset requirement, and obtains the current examinee’s Chinese proficiency level when the accuracy rate does not meet the preset requirement, and continues to push the diagnostic module test question bank of the current Chinese proficiency level to the client.
[0042] Specifically, the server pushes an adaptive module question bank of a corresponding level on the basis of the examinee information provided by the client, such as gender, age, educational background, language background, or the like. The server dynamically selects test questions of appropriate difficulty on the basis of the examinee’s responses, to assess the examinee’s Chinese proficiency level across multiple dimensions, including pronunciation, listening, reading, writing, or the like. If the accuracy rate of the examinee’s responses meets a preset requirement, the server will push the adaptive module question bank of a higher level to the client based on the examinee’s performance, to further assess the examinee’s ability. If the accuracy rate of the examinee’s responses does not meet the preset requirement, the server will obtain the current examinee’s Chinese proficiency level and continue to push a diagnostic module test question bank of the current Chinese proficiency level to the client.
[0043] Herein, the test question information database is divided into an adaptive module question bank and a diagnostic module test question bank, the current examinee’s Chinese proficiency level is determined based on the adaptive module question bank, and the current examinee’s Chinese abilities under the current Chinese proficiency level—including listening, speaking, reading, and writing abilities—are determined based on the diagnostic module question bank.
[0044] Specifically, the present disclosure provides a method for applying a multidimensional adaptive model in Chinese language testing, to achieve the consolidation of scores across multiple dimensions and a precise analysis of multiple sub-skills.
[0045] The test question information database is divided into an adaptive module question bank and a diagnostic module question bank, where the adaptive module question bank contains questions across multiple dimensions, such as listening, speaking, reading, writing, etc., and the diagnostic module question bank contains questions used to perform a proficiency diagnosis on an examinee.
[0046] On the basis of the adaptive module question bank, the current examinee’s Chinese proficiency level is determined by an adaptive algorithm. The adaptive algorithm automatically adjusts a difficulty coefficient based on the examinee’s responses to achieve an optimal testing result.
[0047] On the basis of the diagnostic module question bank, the Chinese abilities under the current Chinese proficiency level are determined, including listening, speaking, reading, and writing abilities. The diagnostic module question bank contains questions across multiple dimensions, corresponding to the questions in the adaptive module question bank.
[0048] The scores across each dimension are combined into a total score to provide a comprehensive evaluation of the examinee’s Chinese abilities.
[0049] By a multidimensional adaptive model, a precise analysis of examinees’ performance across various sub-skills is achieved to identify their advantages and disadvantages. For imperfect sub-skills, related learning recommendations and resources can be provided to help examinees further enhance their abilities. Meantime, it is possible to comprehensively assess the examinees’ Chinese abilities while providing the precise analysis and recommendations, enabling the examinees to engage in targeted learning to address their disadvantages, thereby improving their Chinese proficiency levels.
[0050] Specifically, the discrimination, test reliability, and validity of each question are analyzed, and the questions that meet the preset requirement are incorporated into the adaptive module question bank.
[0051] More specifically, the test question information database is screened based on criteria including the validity, the discrimination D-value, an item-total score correlation coefficient, and Cronbach’s alpha as an overall reliability coefficient to obtain an adaptive module question bank that meets the preset requirement, and the screened adaptive module question bank is used to determine the current examinee’s Chinese proficiency level based on a Chinese diagnostic system;
[0052] where the validity is a statistical indicator that evaluates the relevance and consistency of the questions with subject knowledge and ability standards;
[0053] the discrimination D-value is a statistical indicator that evaluates the discrimination of individual observation items in a measurement tool;
[0054] the item-total score correlation coefficient is an indicator used to measure the relationship between individual questions and the total score in an exam or test;
[0055] Cronbach’s alpha as the reliability coefficient is a statistical indicator that assess the internal consistency of measurements.
[0056] Specifically, the validity involves analyzing test question response and test paper reliability.
[0057] Specifically, the discrimination D-value is calculated as follows.
[0058] Step S11.1: Sort test-takers by their test scores from highest to lowest, and select the top n test-takers as the highest-performing ones and the bottom n test-takers as the lowest-performing ones;
[0059] Step SI 1.2: Calculate average scores of the highest-performing test-takers and the lowest-performing test-takers respectively, and calculate a difference between the two;
[0060] Step S11.3: Calculate a standard deviation for a certain question;
[0061] Step Sil.4: Divide the calculated difference between the two by the standard deviation to obtain the discrimination D-value;
[0062] Step Sil.5: If the discrimination D-value is less than a preset value, the current question is deleted.
[0063] Specifically, the item-total score correlation coefficient is calculated as follows.
[0064] Step S12.1: Standardize each question score and the total score for each test-taker;
[0065] Step S12.2: Calculate a covariance between each question score and the total score;
[0066] Step S12.3: Calculate standard deviations of each question score and the total score;
[0067] Step S12.4: Divide the calculated covariance by the product of the calculated standard deviations to obtain the item-total score correlation coefficient;
[0068] Step S12.5: If the correlation coefficient of a certain item is lower than a preset value, the current item is deleted.
[0069] Specifically, Cronbach’s alpha as the reliability coefficient is calculated as follows: ( 2 A
[0070] ^ = --k-1 cr
[0071] where K represents the number of test questions, g2, represents the variance of scores for each test question, and o2 represents the total variance of scores for the entire exam.
[0072] When the Cronbach’s alpha coefficient is less than a preset value, the test question information database is re-screened based on the validity, the discrimination D-value and the item-total score correlation coefficient, until the Cronbach’s alpha coefficient is greater than the preset value.
[0073] The screened adaptive module question bank is used to determine the current examinee’s Chinese proficiency level based on the Chinese diagnosis system as follow.
[0074] S22.1: On the basis of questions in the adaptive module question bank, perform a natural language processing on the questions to extract keywords and sentences;
[0075] Step S22.2: Use a prompt project to conduct quality audit on the questions, and determine difficulty levels and quality of the questions;
[0076] Step S22.3: Intelligently grade the questions according to the audit results, and assign the corresponding difficulty levels to the questions.
[0077] Specifically, a natural language processing technology is used in conjunction with language corpus, a prompt vocabulary containing keywords and prompt sentences is generated. In the stage of question quality audit, the keywords and sentences in the questions are automatically analyzed and matched with the prompt vocabulary to determine the difficulty levels and quality of the questions. In the stage of intelligent grading, the corresponding difficulty levels are assigned to the questions according to the audit results, so that the number of questions at each difficulty level is evenly distributed to ensure that the difficulty levels of the questions are evenly distributed.
[0078] On the basis of the adaptive module question bank and the prompt vocabulary,by using Chinese language ability growth curve, value-added evaluation and vertical growth model, each question is intelligently graded to determine its language proficiency level and corresponding score value. The score values accross multiple dimensions are weighted and summed, and the total score of the examinee in terms of Chinese language abilities is obtained. At the same time, through an accurate analysis for multiple sub-abilities, detailed error analysis, evaluation report and personalized learning suggestions are provided for examinees. The present disclosure realizes the quality audit and intelligent grading of the adaptive module question bank, provides a more scientific and accurate evaluation system for Chinese testing, and has higher practical value and broad application prospects.
[0079] At the same time, the server pushes the classical measurement theory model and the item response theory model to the client based on the adaptive module question bank, the client feeds back the corresponding responses to the server, and the server determines the current examinee’s Chinese proficiency level based on the responses. Under the framework of the classical measurement theory, each difficulty level matches the course objective content, the error rate is calculated relatively independently for each level, and the levels are upgraded step by step. Under the framework of the item response theory model, the difficulties of all questions are treated equally to form a unified question bank. In the process of specific responses, there will be repetitions of questions of different difficulty, and the level will be determined according to the final score results. Two sets of computing logics are deployed in different link portals and called according to the needs of users.
[0080] Specifically, the server counts scores and erroneous options from examinees’ responses, performs an error attribution, and generates a corresponding report for the Chinese proficiency level, real-time ranking, and ability description.
[0081] Specifically, in the process of multi-dimensional adaptive testing, the server counts score and erroneous options from examinees’ responses, and perform an error attribution on the erroneous options, so as to evaluate the examinees' Chinese proficiency levels across different dimensions. By analyzing the examinees' performance in answering questions, the server generates a corresponding Chinese proficiency level report, which describes the examinees’ performance across different Chinese dimensions in detail. At the same time, the server generates a real-time ranking based on the examinees’ performance in the test, allowing the examinees to assess and compare their own results. The server also generates an ability description report that provides examinees and teachers with detailed assessment results and recommendations, helping them develop more targeted learning plans and teaching strategies. To enhance the accuracy and reliability of assessment, the server performs a comprehensive analysis and data mining on test data, and employ various statistical methods and data mining techniques to find deeper patterns and features. Through these analyses and data mining, the server can provide a more accurate, comprehensive, and personalized assessment on the examinees’ Chinese abilities, along with more targeted learning and teaching recommendations.
[0082] Specifically, attribute tags are added to question stems and options in the test question information database of the Chinese proficiency diagnostic system. While diagnosing the Chinese abilities, the causes of errors are analyzed based on these attribute tags.
[0083] More specifically, vocabulary types and error types related to vocabulary proficiency indicators are defined according to theories established by the Chinese linguistics community, and the defined vocabulary types and error types related to vocabulary proficiency indicators are incorporated into a linguistic attribute tag library. The question stems and options in the test question information database are each annotated with multidimensional linguistic attribute tags based on the linguistic attribute tag library, thereby establishing a test question information database with the linguistic attribute tags.
[0084] The vocabulary types include:
[0085] academic and non-academic vocabularies classified by usage context, as indicated by the attribute tag “Purpose”;
[0086] content words and function words classified by part of speech, where the content words include nouns, verbs, adjectives, and measure words; the function words include adverbs, prepositions, and conjunctions, as indicated by the attribute tag “POS (Part of Speech)”.
[0087] The error types related to vocabulary proficiency indicators include: phonological error types, morphological error types, and semantic error types.
[0088] The phonological error types include: tone errors, spelling rule errors, and confusion with homophones, as indicated by the attribute tag “ErroRecommend (Erroneous Option Diagnosis Description)”.
[0089] The morphological error types include: stroke order errors, structural errors, and missing or extra strokes, as indicated by the attribute tag “ErroRecommend (Erroneous Option Diagnosis Description)”;
[0090] The semantic error types include: confusion of synonyms, confusion of polysemous words, and misuse of words, as indicated by the attribute tags “Antonyms,” “Cognates,” “Polysemy,” “Synonyms,” “Collocations,” “Categories,” and “Optimal.”
[0091] More specifically, the corresponding English and Chinese terms in the attribute tags are as follows.
[0092] Item: number; N: number of samples; P: difficulty value; D: discrimination coefficient; Corr: I-T correlation; Dalpha: delete alpha; DifClass: native language item function difference; DifCurriculum: curriculum item function difference; DifGender: gender item function difference; ItemSerialNumber: item bank serial number; SkillStandards: skill standards; Vocabulary: target vocabulary; POS: part of speech; Level: word level; Instructions: item type instructions; SubAbility: vocabulary ability type; StemFormat: stem format; OptionFormat: option format; Stem: stem; RightOptions: right options; ErroOptions: erroneous options; Purpose: purpose; RightRecommend: right option diagnostic description; ErroRecommend: erroneous option diagnostic description.
[0093] Radar charts for each vocabulary skill are generated based on the scores of item attribute tags, and detailed diagnostic reports for each vocabulary skill are generated based on the selected options in the questions and their corresponding attribute tags.
[0094] The system according to the present disclosure utilizes an artificial intelligence technology to analyze the examinees’ responses, enabling a precise identification of the examinees’ weaknesses and error types in Chinese abilities, thereby providing corresponding instructional recommendations and guidance. Specifically, by analyzing the examinees’ responses, it is possible to automatically diagnose error types of the examinees in Chinese abilities—including grammatical, lexical, and semantic errors—and perform an attribution analysis on these errors to determine the causes of errors. Meantime, this technology can analyze the examinees’ scores across different knowledge points and skills to diagnose weaknesses in their Chinese abilities and provide corresponding instructional recommendations and guidance.
[0095] The implementation of this technology relies on artificial intelligence techniques such as deep learning and natural language processing. By establishing a large-scale Chinese corpus and a Chinese proficiency diagnostic model, and utilizing a deep learning technology to analyze the examinees’ responses, it is possible to quickly and accurately determine the error types and their causes. Meantime, this technology can also analyze the examinees’ scores across different knowledge points and skills. By analyzing the scores, it is possible to diagnose weaknesses of the examinees in their Chinese abilities and provide corresponding instructional recommendations and guidance.
[0096] Specifically, scores and erroneous options from the responses of all examinees are counted and score reports organized in units of class or the whole school are automatically generated.
[0097] More specifically, by combining criterion-referenced and norm-referenced approaches, test results are scored and graded, and score analyses and reports are then provided separately based on personalized feedbacks from school administrators, teachers, and students.
[0098] The criterion-referenced approach involves comparing the test results with a fixed standard or criterion to determine the scores and grades.
[0099] The norm-referenced approach involves comparing the test results with those of a preset number of examinees to determine the position and level of the test results within the overall distribution.
[00100] To address school administrators’ needs, the report provides an overall analysis of student distribution. This includes the distribution of students across different Chinese proficiency levels, pass rates, and excellence rates by grade and class, helping the school administrators understand the overall performance of Chinese language instruction and providing a reference for formulating educational strategies.
[00101] To address teachers’ instructional needs, the report provides comparisons and follow-up teaching recommendations based on students’ background information. This includes analyzing the performance of native and non-native Chinese speakers across various Chinese language concepts and skills, comparing performance differences among students, identifying issues in teaching, and offering follow-up teaching recommendations to help teachers improve the quality of instruction.
[00102] To address individual students’ needs, the report provides personalized diagnostic feedback and follow-up learning recommendations. This includes analyzing students’ performance on different error types based on their scores, establishing longitudinal growth curves, and developing personalized learning plans and recommendations to help students improve their learning outcomes and facilitate effective communication between family and school.
[00103] The system according to the present disclosure offers numerous advantages in Chinese language testing. It can not only enhance the accuracy and efficiency of testing but also enable personalized assessments based on the test-taker’ background information and testing objectives, thereby achieving a more comprehensive and effective measurement of learners’ Chinese proficiency levels.
[00104] Embodiment 2
[00105] The embodiment 2 is a preferred example of the embodiment 1.
[00106] A Chinese examination system suitable for learners with diverse backgrounds according to the present disclosure comprises:
[00107] Step 1 : Formulate a vocabulary selection specification document to select vocabularies for testing;
[00108] Step 2: Formulate a specification document for vocabulary test types, stems, and options, and create questions that meet testing standards;
[00109] Step 3 : Organize a standardized test involving over 4,000 participants, conduct a quality analysis of each question, and assign a difficulty value to each question;
[00110] Step 4: Incorporate questions that pass the quality analysis into the adaptive system question bank;
[00111] Step 5: Develop an adaptive system algorithm for the Chinese language exam;
[00112] Step 6: The adaptive system algorithm automatically determines the difficulty level of the next test question based on the examinee’s current responses and pushes it to the examinee. The system stops pushing questions of the next difficulty level until the examinee’s error rate stabilizes at a certain level, and then assigns a Chinese proficiency level to the examinee.
[00113] Specifically, Step 2 comprises:
[00114] Question Development and Review Phase: Three researchers participated in question development and review. In the first round, the researchers adapted the question-writing principles proposed by Ayano Tsubota and Hidetoki Ishii (2020) to suit the characteristics of the Chinese language, with two researchers each creating two subsets of test questions. In the second round, the researchers cross-checked their work to ensure consistency in the annotation of question attributes (part of speech, purpose, etc.), and received confirmation and feedback from the third researcher. In the third round, senior Chinese language teachers outside the development team were invited to provide feedback and confirmation on the questions.
[00115] Specifically, Step 3 comprises the following sub-steps:
[00116] Step 3.1. Use a quality analysis tool to analyze the Cronbach s consistency coefficient of test questions, the discrimination D-value for each question, and the item-total score correlation, to determine that the vocabulary test is unidimensional and suitable for application in adaptive system algorithms;
[00117] Step 3.2: Conduct criterion-related validity analysis using the same sample’s performance on external authoritative exams to determine that the vocabulary test can represent students’ overall Chinese proficiency levels, and that students’ proficiency levels assessed by the adaptive algorithm can serve as the overall Chinese proficiency levels.
[00118] Criterion-related validity refers to the association between scores on the Chinese vocabulary test and results from other mainstream Chinese proficiency assessments. This study used the reading scores from the Hong Kong campus participants in the “Smart Reading” program developed by Yao-Ting et al. (2014) as a reference. The results of four subtests are as follows. Chinese Examination System of The Present Disclosure External Validity Object Validity Values Levels 1-2 Test SmartReading Elementary School 0.54 (n=128) Levels 3-4 Test SmartReading Elementary School 0.67 (n=62) Levels 5-6 Test SmartReading Elementary 0.72 (n=148) School Levels 7-8 Test SmartReading Middle School (First Language) 0.69 (n=67)
[00119] Overall, the result of the Chinese vocabulary test can predict scores of the students on the reading test to some extent, which are consistent with the findings of Cheng and Matthews (2018) and Schmitt (2014).
[00120] The Cronbach’s consistency coefficient, which is a measure of the test reliability and also known as the internal consistency coefficient, is a statistical indicator used to measure the reliability of a scale or questionnaire in terms of internal consistency. Specifically, it is a method for measuring the correlation among multiple observation items (such as questions in the questionnaire) to determine whether they measure the same concept or dimension.
[00121] The Discrimination D-value is also known as a “high-low analysis,” and the Discrimination D-value is a statistical indicator used to assess the discrimination of individual observation items in a measurement tool. Specifically, it is used to measure a question’s ability to distinguish populations with different proficiency levels—that is, whether a question can identify those with high scores and those with low scores.
[00122] The item-total score correlation refers to the correlation between the score of each observation item (i.e., question) in a certain measurement tool and the total score of the entire measurement tool. This correlation can help us understand the extent to which each question contributes to the total score, as well as whether certain questions have a weaker correlation with the total score and may therefore require revision or removal.
[00123] Step 3 further involves using the Chinese vocabulary test as an officially organized examination. The curricula used include the Chinese language curriculum for compulsory education in China, the local Chinese curriculum in Hong Kong, and international curricula. The student population includes native and non-native speakers, as well as those learning Chinese as a first, second, or third language, etc., making it an appropriate research sample for developing a Chinese vocabulary test tailored to learners with diverse backgrounds. A total of 4,464 students’ responses data are included in the study.
[00124] Since school-organized exams that students engage in are less likely to guess, and to maintain consistency with previous research, the present disclosure employs a two-parameter model, which calculates only the difficulty and discrimination of questions. Equivalence processing was performed three times using the Mean / Sigma method in linear transformation.
[00125] The indicator to measure the test reliability can refer to the Cronbach’s consistency coefficient. The high quality of each question is a prerequisite for a high reliability of the whole test. Measures of question quality include the discrimination D-value and the item-total score correlation of the questions.
[00126] The standard requires a D-value of no less than 0.2 and an item-total score correlation of no less than 0.3, and questions that fail to meet these criteria are deleted, as shown in FIG. 8.
[00127] Unqualified questions tend to exhibit an excessively high or low accuracy rate, meaning that regardless of a learner’s proficiency level, he / she will consistently answer these items rightly (or wrongly). This will lead to a ceiling effect or a floor effect. If the unqualified questions are present in the adaptive algorithm process, it becomes difficult to accurately assess the learner’s true level. In other words, this will either reduce the accuracy of the adaptive algorithm or require an increase in the number and allotted time of test questions to maintain the same accuracy, which can have an adverse consequence.
[00128] The system according to the present disclosure is suitable for Chinese learners with diverse backgrounds, including both native and non-native speakers. It intelligently pushes questions, offers timely feedback, and automatically and promptly generates various score reports tailored for different audiences.
[00129] The Chinese vocabulary test question bank provided by the present disclosure can accurately assess learners’ Chinese levels, as it includes Chinese vocabularies accross various difficulty and proficiency levels, enabling a comprehensive and accurate evaluation of learners’ Chinese levels. This is crucial for both learners’ self-evaluation and teachers’ instructional evaluation. Furthermore, the establishment of the Chinese vocabulary test question bank provides a foundation for theoretical research and teaching practices that rely on Chinese proficiency as a benchmark. Researchers can utilize this test question bank to conduct various studies on Chinese language assessment, thereby advancing the development of Chinese assessment theory. Meantime, teachers can also use the content of the test question bank to develop lesson plans and instructional strategies, thereby better facilitating students’ Chinese language learning. Finally, by exploring the relationship between the difficulty of questions and the vocabulary levels, the establishment of the Chinese vocabulary test question bank can, to a certain extent, deepen our understanding of Chinese language assessment and instruction. Researchers can analyze learners’ learning patterns and challenges based on data from the test question bank, thereby providing more scientific and effective methods and strategies for Chinese language instruction. At the same time, teachers can also use the test question bank to better understand students’ learning progress and further improve teaching effectiveness.
[00130] The adaptive examination system provided by the present disclosure can conduct personalized diagnosis and assessment of students’ language proficiency, learning styles, and subject knowledge under complex learning conditions. It provides teachers with more accurate insights into students’ abilities and progress, enabling them to better develop personalized teaching plans and improve teaching effectiveness. At the same time, the system can automatically adjust question difficulty based on students’ responses, giving each student an opportunity to reach a higher proficiency level. This stimulates their motivation and self-confidence, enhancing their enthusiasm and initiative in learning. The adaptive examination system offers students diverse learning resources and assessment methods, such as audio, text-and-image formats, and interactive assessments, thereby increasing students’ interest and improving learning outcomes. The adaptive examination system intelligently designs the difficulty and types of test questions based on students’ abilities and proficiency levels, ensuring that every examinee can take an exam in a relatively fair environment and avoiding the impact of individual differences and subjectivity on test results in traditional exams. Finally, the adaptive examination system can intelligently collect and analyze students’ learning data and results, providing educational assessments and feedback. This offers a strong support for the continuous improvement and advancement of Chinese language education, and also facilitates the system’s widespread adoption and application, thereby providing a robust support for the modernization and internationalization of Chinese language education.
[00131] The timely, visualized score reports of the system according to the the present disclosure help to monitor students’ learning progress. By evaluating students’ academic performance, educators can assess their mastery of various knowledge points and skills. At the same time, timely, visualized score feedback can allow students and parents to understand the students’ learning progress, thereby enabling them to take corresponding measures to improve learning outcomes. Meantime, it promotes personalized instruction: assessed and timely, visualized score feedback provide foundational data for personalized teaching. Teachers can develop corresponding teaching plans and strategies based on students’ scores to meet the needs of different students. At the same time, students and parents can also use the score feedback to identify students’ advantages and disadvantages, and choose appropriate learning and tutoring methods, to foster student motivation: the timely, visualized score feedback can motivate students to learn. Students can analyze and compare their own grades to identify their advantages and disadvantages, thereby developing more targeted study plans and maintaining a positive attitude toward learning throughout the process. Finally, by fostering home-school collaboration, the timely, visualized score feedback can foster home-school collaboration. Parents can use the score feedback to understand their child’s academic progress, communicate with teachers to identify learning challenges, and seek advice and assistance from teachers. The system according to the present disclosure also incorporates automated AI technology that analyzes students’ academic performance and responses to automatically generate personalized learning recommendations and instructional strategies, thus helping students better master key concepts and skills. At the same time, teachers can also use these generated recommendations and strategies to provide more targeted guidance and tutoring, thereby improving students’ learning outcomes. This automated technology can significantly reduce the workload of teachers and improve teaching efficiency, while providing students and parents with more comprehensive and accurate learning assessments and feedback, thereby fostering home-school collaboration and jointly advancing students’ academic development.
[00132] Those skilled in the art should appreciate that, in addition to implementing the system, apparatus, and their respective modules provided by the present disclosure in the form of purely computer-readable program codes, the same functionality can be achieved by logically programming the method steps to implement the system, apparatus, and their respective modules using logic gates, switches, application-specific integrated circuits (ASICs), programmable logic controllers (PLCs), and embedded microcontrollers. Therefore, the system, apparatus, and their respective modules provided by the present disclosure may be regarded as a hardware component, and the modules contained therein for implementing various programs may also be viewed as structures within the hardware component. Furthermore, the modules for implementing various functions may also be regarded as either a software program that implements the method or structures within the hardware component.
[00133] The above describes specific embodiments of the present disclosure. It should be understood that the present disclosure is not limited to the specific embodiments described above. Those skilled in the art could make various changes or modifications within the scope of the appended claims without departing from the spirit and scope of the present invention. The embodiments of the present disclosure and features thereof may be arbitrarily combined with one another without conflict.
Claims
1. A Chinese examination system suitable for learners with diverse backgrounds, comprising: a server, a client, and a data transmission network;wherein the client is a terminal device for a Chinese examination;the server provides a database service for the Chinese examination and comprises a test question information database, the test question information database comprising an adaptive module question bank and a diagnostic module question bank;the data transmission network is connected to the server and the client and is used for guaranteeing data transmission between the server and the client;when the Chinese examination begins, the server pushes the adaptive module question bank of a corresponding level on the basis of examinee information provided by the client; the client transmits the corresponding responses to the server, the server determines the difficulty level of the next round of test questions on the basis of the responses provided by the client and pushes the next round of test questions to the client, until the error rate of the responses of the adaptive module question bank of the corresponding level meets a preset condition, and obtains the current examinee’s Chinese proficiency level; the server pushes the diagnostic module question bank corresponding to the current examinee’s level to the client, the client feeds back the corresponding responses to the server, and the server performs a diagnosis assessment under the current Chinese proficiency level on the basis of the responses.
2. The Chinese examination system suitable for learners with diverse backgrounds according to claim 1, characterized in that, the server pushes a test question version on the basis of a font for examination selected by the client.
3. The Chinese examination system suitable for learners with diverse backgrounds according to claim 1, characterized in that, the server pushes the adaptive module question bank of the corresponding level on the basis of the examinee information provided by the client, the server determines the responses provided by the client, pushes the adaptive module question bank of a higher level to the client when the accuracy rate meets a preset requirement, and obtains the current examinee’s Chinese proficiency level when the accuracy rate does not meetthe preset requirement, and continues to push the diagnostic module test question bank of the current Chinese proficiency level to the client.
4. The Chinese examination system suitable for learners with diverse backgrounds according to claim 1, characterized in that, the server counts scores and erroneous options from examinees’ responses, performs an error attribution, and generates a corresponding report for the Chinese proficiency level, real-time ranking, and ability description.
5. The Chinese examination system suitable for learners with diverse backgrounds according to claim 1, characterized in that, scores and erroneous options from examinees’ responses of all examinees are counted and score reports organized in units of class or the whole school are automatically generated.
6. A Chinese examination method suitable for learners with diverse backgrounds, comprising:Step SI: a server pushes an adaptive module question bank of a corresponding level on the basis of examinee information provided by a client;Step S2: the client transmits corresponding responses to the server, the server determines the difficulty level of the next round of test questions on the basis of the responses provided by the client and pushes the next round of test questions to the client, until the error rate of the responses of the adaptive module question bank of the corresponding level meets a preset condition, and obtains the current examinee’s Chinese proficiency level;Step S3: the server pushes a diagnostic module question bank corresponding to the current examinee’s level to the client, the client feeds back the corresponding responses to the server, and the server performs a diagnosis assessment under the current Chinese proficiency level on the basis of the responses.
7. The Chinese examination method suitable for learners with diverse backgrounds according to claim 6, characterized in that, the server pushes a test question version on the basis of a font for examination selected by the client.
8. The Chinese examination method suitable for learners with diverse backgrounds according to claim 6, characterized in that, the server pushes the adaptive module question bank of the corresponding level on the basis of the examinee information provided by the client, the server determines the responses provided by the client, pushes the adaptive module question bank of a higher level to the client when the accuracy rate meets a preset requirement, obtains the current examinee’s Chinese proficiency level when the accuracy rate does not meet the preset requirement, and continues to push the diagnostic module test question bank of the current Chinese proficiency level to the client.
9. The Chinese examination method suitable for learners with diverse backgrounds according to claim 6, characterized in that, the server counts scores and erroneous options from examinees’ responses, automatically performs an error attribution, and generates a corresponding report for the Chinese proficiency level, real-time ranking, and ability description.
10. The Chinese examination method suitable for learners with diverse backgrounds according to claim 6, characterized in that, scores and erroneous options from responses of all examinees are counted and score reports organized in units of class or the whole school are automatically generated.
Citation Information
Patent Citations
Language competence test method and system
CN105989555A
Terminal assessment method and device
CN106652620A
Language adaptive hierarchical learning method and device
CN111932415A
Learning ability level diagnosis system
CN112614030A
Chinese examination system suitable for learners with multiple backgrounds and setting of actual operation and feedback of Chinese examination system
CN117496775A
