Self-comparison type occupational interest evaluation method and online adaptability test system

Abstract

Existing occupational interest assessment methods and systems are low in efficiency and insufficient in resistance to response deviation. The invention provides a multi-dimensional forced selection method based on a self-comparison mode. The multi-dimensional forced selection method based on the self-comparison mode comprises the steps that a multi-dimensional forced selection method is provided, and the multi-dimensional forced selection method based on the self-comparison mode is provided; the invention relates to a method and an online self-adaptive test system for an MFC question, which are used for evaluating the occupational interest level of a user. According to the method, corresponding questions can be selected and tested according to the specific occupational interest level of each user, and meanwhile, the exposure rates of the questions are controlled, so that an efficient and stable solution is provided for occupational interest assessment. By using MFC questions, the method effectively reduces common reaction deviation and false possibility in occupational interest assessment. Through the online adaptive test system, the method can realize real-time, efficient and accurate evaluation of the occupational interest of the user.

Description

[0001] This application claims priority to the Hong Kong Patent No. 32024095423.2, filed on August 12, 2024, the entire contents of which are incorporated herein by reference. Technical Field

[0002] The non-limiting and exemplary embodiments of this disclosure relate to the general field of information technology, and specifically to a method and online adaptability testing system for self-comparison assessment of a user's level of career interest. Background Technology

[0003] This section provides background information for a better understanding of this disclosure. Therefore, the statements in this section should be interpreted in this context and should not be construed as an admission that something is prior art or not.

[0004] 1-Introduction to Career Interest Assessment

[0005] Career interest assessment is a systematic process that helps individuals identify their preferences and inclinations for different types of work. It guides people to make informed career choices by matching personal interests with suitable job roles. In the field of career planning, it plays a crucial role in cultivating self-awareness, ensuring that individuals can pursue careers that truly align with their interests.

[0006] 2. Overview of Holland's Occupational Interests Assessment

[0007] Dr. John L. Holland's theory of vocational interests is a crucial theory in the field of career counseling. This theory posits that vocational interests can generally be categorized into six types: Realistic (R), Investigative (I), Artistic (A), Social (S), Enterprising (E), and Conventional (C). Each vocational type is associated with a unique set of job characteristics and work environments. The Holland Vocational Interest Assessment, developed based on this theory, has a long history and is widely recognized for its effectiveness in career guidance.

[0008] Individuals participating in the Holland Occupational Interests Assessment typically complete a self-report paper-and-pencil or online questionnaire to assess their preferences for various activities and work styles. The questionnaire results generate a profile showing their primary areas of interest. This profile can be compared to interest codes for different occupations to identify an individual's potential career paths.

[0009] Examples of online tests include the Holland Code (RIASEC) test (https: / / openpsychometrics.org / tests / RIASEC / ) and the Holland Code Career Test (https: / / www.truity.com / test / holland-code-career-test). It's important to note that although these tests are administered online, they provide every test-taker with the same questions and require them to complete them within a fixed timeframe. Essentially, they are digital versions of traditional paper-and-pencil career interest tests. Unlike traditional paper-and-pencil tests, these online tests offer greater convenience because they do not require users to take them at a specific time and place. However, they do not improve testing efficiency, as all users still have to answer the same questions and spend almost the same amount of time.

[0010] Furthermore, these existing testing systems typically use Likert-type questions. Likert-type questions are well-known to be susceptible to response biases. Therefore, when Likert-type questions are used for career interest assessments (such as pre-employment screening), user response biases are very common. For example, they tend to choose answers that align with societal expectations rather than responding based on their own true abilities.

[0011] In summary, career interest assessments like the Holland Occupational Themes Assessment provide individuals with valuable tools. Through these assessments, individuals can plan their career paths, ensuring that their work is not merely a means of making a living, but also a source of satisfaction and fulfillment. Summary of the Invention

[0012] As stated above, the inventors of this disclosure have found that existing methods and systems for assessing career interests are neither efficient nor effective in resisting response bias. Benefiting from the rapid development of computer technology and psychometric theory, the inventors of this disclosure have devised a solution to address or at least mitigate these problems. The main objectives of this solution include:

[0013] 1. To conduct career interest assessments more efficiently and conveniently.

[0014] 2. Effectively prevents response bias in career interest assessments, and

[0015] 3. Develop an online system for assessing career interests that possesses the advantages mentioned above.

[0016] The first aspect of this approach is achieved through Computerized Adaptive Testing (CAT) technology. CAT utilizes computer technology and advanced psychometric models and algorithms to enable online assessment of adaptability. The CAT process includes the following steps: (1) constructing a question bank containing hundreds of high-quality items and calibrating (i.e., pre-testing) these items on a representative sample of users; (2) setting an initial career interest level (e.g., 0) for each user to initiate the CAT process; (3) selecting items from the question bank suitable for each user's career interest level while controlling the exposure of the items; (4) updating the estimate of the user's career interest level based on the user's answers to the selected items; (5) determining whether the number of items administered has reached the pre-set test length; if not, repeating steps 3 and 4; and (6) outputting the career interest level as the assessment result. Steps 1 and 2 are performed before the items are administered, while the remaining steps are the measurement process for individual users. Detailed steps are described in the "Detailed Implementation" section.

[0017] This disclosure provides an efficient method for assessing career interests because, for each user, only questions that match their level of career interest are administered, while questions that are unsuitable for their level are not. By introducing a random component in the question selection process, this disclosure effectively controls the exposure rate of questions, thereby avoiding overexposure or potential leakage and ensuring the security of the testing process.

[0018] By leveraging internet technology, this disclosure further simplifies the CAT process, promoting efficient administration and immediate score reporting. Therefore, it allows users to test their career interests anytime, anywhere using the Holland Occupational Interests Scale, significantly reducing testing time and providing immediate results.

[0019] The second aspect of this approach is achieved through the use of Multidimensional Forced-Choice (MFC) questions in career interest assessment. MFC questions typically consist of two statements that are similar in social expectation, each measuring one of six career types. Furthermore, K options reflect K preference categories. Users choose one option from these K options when taking the test (hence the name "forced-choice"). As mentioned above, unlike Likert-type questions used in existing systems, MFC questions effectively reduce response bias or fraud. For brevity, unless otherwise stated, the term "question" below specifically refers to an "MFC question."

[0020] The third aspect of this approach is achieved through an online system for administering a self-comparison-based Holland Occupational Interests Assessment. In this system, only items appropriate to the user's level of occupational interest are selected and administered to that user. This system includes a suite of computer program products, such as instructions to execute the algorithms and steps described in the first aspect of the approach, a processor to execute the computer program, and memory and other computer-readable media to store the instructions, user responses collected during the test, and other information. All of these methods are implemented on a secure, highly scalable, cost-effective, and manageable platform: the Self-Comparison Holland Occupational Interests Online Adaptability Test System (IHCI-CAT). Attached Figure Description

[0021] The above and other aspects, features, and advantages of this disclosure will be described in detail below with reference to the accompanying drawings. Similar reference numerals or letters are used to denote similar or equivalent elements. The drawings are intended to better explain embodiments of this disclosure and are not necessarily drawn to scale, wherein:

[0022] Figure 1 The method flowchart of this disclosure is described;

[0023] Figure 2 This publication describes example MFC problems;

[0024] Figure 3 The development of this publicly available test item bank is described;

[0025] Figure 4 A sample report generated by the online testing system disclosed herein is described;

[0026] Figure 5 A schematic block diagram of the system disclosed herein is described;

[0027] Figure 6 Another schematic block diagram of the system disclosed herein is described. Detailed Implementation

[0028] The embodiments described herein will be described more fully below with reference to the accompanying drawings. However, the embodiments herein may be implemented in many different forms and should not be construed as limiting the scope of the appended claims.

[0029] The terminology used herein is for describing particular embodiments only and is not intended to be limiting. The singular forms “-” (“a”, “an”) and “the” as used herein also include the plural forms unless the context clearly indicates otherwise. Furthermore, the terms “comprises”, “comprising”, and / or “includes” (“including”) as used herein specify the presence of the stated features, integers, steps, operations, elements, and / or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or combinations thereof.

[0030] Furthermore, the use of ordinal numbers (such as "first", "second", "third" etc.) to modify elements in this article does not indicate the priority, priority, or order of one element relative to another element, nor does it indicate the temporal order of the execution of method actions. Rather, it is used only as a label to distinguish one element with a specific name from another element with the same name (but using ordinal numbers).

[0031] Unless otherwise defined, all terms used herein (including technical and scientific terms) shall have the same meaning as commonly understood. Furthermore, terms used herein shall be interpreted as consistent with their meaning in the context of this specification and the relevant field, and shall not be construed as having an idealized or overly formal meaning unless expressly defined herein.

[0032] Figure 1 A flowchart of a method 100 for self-comparative assessment of a user's career interest level is presented, wherein the career interest level is composed of the user's degree of interest in various career types. Method 100 may include the following steps: Step 101, establishing a question bank containing hundreds of high-quality MFC questions to measure six career types, and calibrating these question parameters on a representative sample. Each MFC question contains two statements that are similar in social expectation and measure two different career types, and K options that reflect K preference categories between the two statements; Step 102: The initial level (i.e., initial value) of each user's career interest is set to 0 to start CAT testing; Step 103: Questions are selected from the question bank based on the user's current career interest level, while controlling the exposure of the questions; Step 104: The career interest level is updated based on the user's answers to the questions selected in Step 103, where each response is one of the K options; Step 105: It is determined whether the number of test questions has reached the preset test length; Step 106: If the termination criterion in Step 105 has been met, the career interest level is output as the result; otherwise, Steps 103 and 104 are repeated.

[0033] The following sections will describe specific embodiments and examples in conjunction with an exemplary assessment system developed by the inventors of this disclosure. This exemplary assessment system (hereinafter referred to as the IHCI-CAT system) is an online computerized adaptability testing system that performs a self-comparison Holland Occupational Interests (IHCI) assessment. The IHCI-CAT system provides an effective and convenient way to assess one's own occupational interests online. Users can... http: / / hkumatlab.testserverhk.com / Access to the system is permitted. However, at this stage, IHCI-CAT is only available to internal staff. It is worth noting that although this exemplary system relates to IHCI assessments, the examples / implementations of this disclosure can also be applied to any other career interest assessment that uses MFC questions.

[0034] As mentioned above, career interest levels consist of a user's level of interest in various career types. For example, in the IHCI-CAT system, a user's career interest level is composed of their degree of interest in six career types: Realistic (R), Investigative (I), Artistic (A), Social (S), Enterprising (E), and Conventional (C).

[0035] Furthermore, as mentioned above, existing vocational interest measurement systems typically use Likert scales. For example, a classic Holland Vocational Interest Scale question might be: "How much do you enjoy social activities? Measure it on a 5-point scale (e.g., dislike at all, rarely like, neutral, quite like, very like)?" When such questions are used in important situations, such as pre-employment screening, job seekers, in an effort to increase their chances of getting the job, are likely to provide false or societally expected responses rather than genuine responses based on their actual interests. This phenomenon is known as response bias.

[0036] In contrast, this disclosure proposes using MFC questions. Since the two statements in an MFC question have a comparable level of social expectation, it can reduce the likelihood of users answering dishonestly. Figure 2 This example question demonstrates an MFC (Multi-Dimensional Forced Choice) question. It contains two statements: "Visit a museum" and "Attend a party." While similar in social expectancy, they measure two different career types. Specifically, "Visit a museum" measures the "Artistic (A)" career type, while "Attend a party" measures the "Social (S)" career type. In other words, these two statements measure career interest levels from different dimensions (i.e., types), reflecting the "multi-dimensional" nature of Multi-Dimensional Forced Choice (MFC) questions.

[0037] Figure 2The example question also shows four options: "I like A much more than B", "I like A more than B", "I like B much more than A", and "I like B more than A". These options reflect four different preference categories (or four different preference intensities) between the two statements. As an answer, the user needs to choose one of the four options. This reflects the "forced choice" nature of Multidimensional Forced Choice Questions (MFC).

[0038] from Figure 2 It can also be seen that the example MFC question includes an instruction: "Please read the two statements and choose one of the four categories based on your preference. For example, if you prefer visiting a museum to attending a party, choose 'I prefer A to B'", to guide the user on how to respond to the question. This instruction is part of the MFC question and can be presented to the user before the question is officially administered, or within each question. The scheme disclosed herein is not limited in this respect.

[0039] CAT technology requires a question bank containing a large number (typically hundreds) of high-quality questions. According to this disclosure, the IHCI-CAT system's question bank contains 540 MFC questions, each with four options, such as... Figure 2 As shown. The question bank may change over time. For example, new questions may be added, overexposed questions may be removed, or outdated questions may be replaced with more suitable ones.

[0040] Figure 3 It demonstrates the formation of the test item bank. Figure 3 In the example, Realistic (R) is measured by S1 to S6, Investigative (I) by S7 to S12, Artistic (A) by S13 to S18, Social (S) by S19 to S24, Enterprising (E) by S25 to S30, and Conventional (C) by S31 to S36. Each MFC question in the question bank is formed by pairing statements of two different career types. For example, {S1, S7}, {S1, S8}, ..., {S1, S12} represent MFC questions generated by pairing the first statement of type R with the six statements of type I; {S1, S13}, {S1, S14}, ..., {S1, S18} represent MFC questions generated by pairing the first statement of type R with the six statements of type A, and so on.

[0041] Therefore, the question bank contains all combinations of any two occupational types, such as Figure 3 The lines connecting the six occupational types are shown in the diagram. This publicly available question bank contains a total of [number missing]. An MFC problem. Although Figure 3Each occupational type is measured by six statements, but this disclosure does not limit the number of statements for each occupational type, nor does it require that the number of statements for each occupational type be the same.

[0042] Of the 540 MFC questions, each occupation type is measured using 180 MFC questions. As mentioned above, each occupation type has six different expressions. Therefore, pairing R with I generates 36 MFC questions, pairing R with A generates another 36 questions, and so on. These 180 MFC questions will be used to measure the R type.

[0043] Because a single statement may be presented to a user multiple times, the user may be affected by the sequence effect. The IHCI-CAT system employs a balancing approach to control these effects. For example, suppose S1 is paired with S8 in question 1 and with S14 in question 2. Therefore, S1 appears twice. To avoid the sequence effect, S1 is presented as the first statement when question 1 {S1, S8} is presented, and as the second statement when question 2 {S14, S1} is presented.

[0044] CAT technology assumes that the parameters of the items in the item bank are known. Therefore, in embodiments, the method includes calibrating the item parameters in the item bank using an Item Response Theory (IRT) model. According to this disclosure, the item parameters in the IHCI-CAT system have been calibrated using an IRT model developed by Qiu and colleagues for MFC items (see the paper "ITEM RESPONSE THEORY MODELS FOR POLY TO MOUS MULTIDIMENSIONAL FORCED-CHOICE ITEMS TO MEASURE CONSTRUCT DIFFERENTIATION" co-authored by Qiu, X.-L., dela Torre, J., Wang, Y.-G., and Wu, J., published in *Educational Measurement: Issues and Practice*, 2024, which is available through... https: / / doi.org / 10.111 / emip.12621 The study (hereinafter referred to as "Qiu et al. 2024") calibrated using responses collected from approximately 300 Hong Kong university students. The item parameters of the model represent the appeal of the statements. Therefore, it can be used to assess which professional activities or career types are more attractive to users.

[0045] In this embodiment, the method includes setting an initial level (i.e., an initial value) for each user's career interests before administering the test. This disclosure follows common practice by setting the initial values ​​of all six career interests to 0. However, it should be noted that the initial values ​​do not necessarily have to be 0, nor do the initial values ​​of the six career interests have to be the same.

[0046] During CAT testing, after previously selected questions have been administered, new questions are selected from the question bank and presented to the user. According to the scheme disclosed herein, question selection is primarily based on the user's level of career interest, which may be an initial value set to zero for the six career types at the start of the assessment, or a level of career interest updated based on the user's responses to previously selected questions. Therefore, in the scheme disclosed herein, the questions administered to each user are different, rather than the same questions administered to all users as in existing systems. Selecting the next question based on the level of career interest ensures that the assessment is personalized to each user's level of interest, thereby achieving adaptive assessment.

[0047] In this embodiment, new topics are selected using criteria that reflect the level of career interest. The statistical information of the topics (i.e., the amount of information they provide for measuring career interest) naturally becomes a basis for this selection process. Generally, the more statistical information a topic provides, the more accurate the estimation of career interest level will be using that topic. Of course, many indicators can represent the statistical information of topics, such as Fisher information (FI), Shannon information, Kullback-Leibler divergence, Akaike information criterion, Bayesian information criterion, and entropy. One or more of these statistical measures can be used to calculate the statistical information of topics according to the specific requirements of the application scenario and software / hardware limitations of this disclosure. This disclosure does not impose any limitations in this regard.

[0048] In this embodiment, question exposure control is a crucial issue in question selection, as it relates to the effective utilization of the question bank and the prevention of overexposure or potential leakage of questions, thereby ensuring the security of the testing process. The solution disclosed herein employs a method combining random and statistical components to control question exposure rates. These two components function at different stages of the test.

[0049] In the initial stages of testing, when users' levels of career interest are unclear, the importance of the information component can be set relatively low, while the importance of the random component can be set relatively high. Especially for the first question, when users' levels of career interest across the six career types are unknown, questions can be randomly selected. This approach allows the testing system to explore users' potential career interests across a broader range.

[0050] As the assessment progresses, users may feel fatigued or lose motivation. At this point, since users have already answered some questions, their level of career interest becomes clearer. Therefore, in the later stages of the assessment, the importance of the informational component can gradually increase, while the importance of the random component can gradually decrease, allowing for the selection of questions that provide more information.

[0051] In summary, in the example IHCI-CAT system, as the test progresses, the influence of the random component on item selection gradually decreases, while the importance of the information component gradually increases. The random component can more effectively utilize the item bank, while the statistical information component ensures the accuracy of the estimation of career interest levels. Combining the two components will help improve the utilization rate of the item bank and the accuracy of the estimation of career interest levels.

[0052] In the embodiment, we utilize the function f v To perform topic selection with exposure control, the function is defined as follows:

[0053]

[0054] Where l is the number of questions administered; T is the number of questions each user will answer (i.e., the pre-set test length); v is a vector containing identifiers of questions not selected during the current test in the question bank; det(I v R represents a vector of the determinants of the FI matrix for each unselected item in v, and is the negative of the expected value of the second derivative of the probability logarithm (the calculation uses updated occupational interest levels, item parameters for each item, and the IRT model developed by Qiu et al. 2024); v This represents each unselected question from a uniform distribution [0, max{det(I v The vector of random numbers generated by )}], where max() represents the maximum value. Here, and These can be viewed as the weights of the random component and the information component, respectively. When problem l′ maximizes f... v At that time, the CAT system selects the topic:

[0055]

[0056] To better understand how the function fv works, we can observe its application in the example IHCI-CAT system. In the IHCI-CAT system, the test length is fixed at 36, meaning each user will answer 36 questions. Therefore, in the function f... v In this context, T = 36. When a user starts the test in the IHCI-CAT system (at which point their occupational interest in the six occupational types is initialized to 0), no questions have been selected yet (i.e., l = 0). Substituting the values ​​of l and T into the function f above... v It becomes Therefore, the IHCI-CAT system will select one with R v The question with the highest value is used as the user's first question. Since the selection relies entirely on randomly generated numbers, this is equivalent to randomly selecting a question from a question bank. As the first question is administered, the user's career interest level is updated.

[0057] Next, we need to choose a second question. In this case, one question has already been chosen (i.e., l = 1). Therefore, The IHCI-CAT system will provide R v Generate another set of random numbers and compute the determinant of FI for each item based on the updated career interest level. Utilize the new R... v and det(I v ), recalculate f v The value, and the IHCI-CAT system selects the value with f v The question with the maximum value is the user's second question. It can be observed that the weight of the random component decreases from 1 to... The weight of the information component increases from 0 to... CAT continues until the last item is selected (i.e., l = 35), at which point the function value is calculated as follows: Similarly, having f v The question with the highest value will be selected as the user's last question.

[0058] Therefore, using f v The IHCI-CAT system selects questions using a function. In the early stages of the assessment, it places more weight on the random component, thus selecting questions with less information. As the test progresses, the IHCI-CAT system gradually prioritizes questions with more information to improve the accuracy of assessing career interest levels.

[0059] In an embodiment, the method further includes controlling the exposure of an expression by removing questions associated with that expression from the question bank according to predetermined criteria. Specifically, during each test administration period, the number of times each expression is presented to the user is recorded and counted. Once an expression exceeds a predetermined maximum number of presentations (e.g., 7 times in the IHCI-CAT system), all questions associated with that expression are removed from the question bank and become unavailable to that user. It is important to note that the removed questions are only inaccessible to the current user; they remain available to other users.

[0060] After a user answers the selected question, their career interest level will be updated based on their response. Of course, different algorithms can be used to update a user's career interest level, and this disclosure does not impose any limitations on this. In this embodiment, the following Newton-Raphson process is used to update the career interest level:

[0061]

[0062] Where l is the number of test items, and θ is the estimate of the six occupational interests; It is a provisional θ estimate from l test items. This is an updated θ estimate. Furthermore, and These are the first and second derivatives of the natural logarithm of the posterior density function of the reaction vector x, which can be calculated as follows:

[0063]

[0064] as well as

[0065]

[0066] Where P k v is the probability of choosing category k (k = 0, 1, 2, ..., K-1). k It is the total score of category k (the sum of scores), and μ and Φ are the mean vector and variance-covariance matrix of the multivariate normal distribution of θ, respectively. The paper “COMPUTERIZED ADAPTIVE TESTING FOR IPSATIVE TESTS WITH MULTIDIMENSIONAL PAIRWISE-COMPARISON ITEMS: ALGORITHMDEVELOPMENT AND APPLICATION” describes this update process. The paper, authored by Qiu, X.-L., de la Torre, J., Ro, S. and Wang, W.-C., was published in Applied Psychological Measurement 46(4), 255-272, 2022, and can be accessed through https: / / doi.org / 10. 1 177 / 01466216221084209 Visited, and is referred to below as "Qiu et al. (2022)".

[0067] After the career interest level is updated, a predefined criterion is used to determine whether the test should be terminated. If the criterion is met, the method terminates the test and outputs the user's career interest level. Otherwise, the test is repeated. Figure 1 Steps 103 and 104 in the above. Of course, different criteria can be used to terminate the test, and the scheme disclosed herein does not limit this.

[0068] The IHCI-CAT system uses a pre-set test length (i.e., 36 questions) as the termination criterion. Therefore, the IHCI-CAT system terminates the assessment once the user has answered 36 questions. Of course, using 36 questions as the termination criterion is based on observation and practice. The IHCI-CAT system allows for different numbers of questions as the termination criterion. In this disclosure, the question selection is primarily adapted to the user's level of occupational interest, so the 36 questions administered to each user are different. It should be noted that the 36 questions administered to each user may not be evenly distributed across the six occupational types. For example, the 36 questions administered by a user may include 5, 6, 8, 7, 6, and 4 questions, measuring R, I, A, E, S, and C, respectively. However, Qiu et al. (2022) found that due to the function f... v The random component and exposure control methods used in the model do not significantly affect the number of questions for each type.

[0069] After the test is completed, the IHCI-CAT system will provide each user with, such as Figure 4 The report presents a final estimate of the user's career interest level. It uses a radar chart to depict the level of career interest in each user's six career interest categories and includes matching occupations and positions. Furthermore, the IHCI-CAT system provides users with the ability to save, print, or directly email the report.

[0070] In this embodiment, the occupational interest level refers to Holland's occupational interest level, which includes the following six occupational types: Realistic (R), Investigative (I), Artistic (A), Social (S), Enterprising (E), and Conventional (C). As described above, the IHCI-CAT system performs a self-comparative assessment of Holland's occupational interest level.

[0071] In this embodiment, each user's career interest level is initialized upon system startup. For example, in the IHCI-CAT system, each user's six career interest levels are set to 0 when they begin the test.

[0072] In this embodiment, an online computerized adaptability testing (CAT) method (e.g., the IHCI-CAT system) is used for evaluation.

[0073] Figure 5A schematic block diagram of an exemplary IHCI-CAT system according to this disclosure is shown. The system includes: a test item calibration module 501 for constructing a test item library and calibrating the parameters of all items on a representative sample using an IRT model; a starting point or entry level module 502 for providing an initial value for the occupational interest level for each user before the test is administered; a test item selection and exposure control module 503 for selecting new items from the test item library suitable for each user's occupational interest level and controlling the exposure of these items; a scoring module 504 for updating the occupational interest level based on the user's answers to the selected items; a termination module 505 for determining whether the number of administered items has reached a pre-set test length. If the termination condition is met, the occupational interest level is output as the evaluation result; otherwise, the operations of modules 503 and 504 are repeated; and a result reporting module 506 for reporting the evaluation results.

[0074] Figure 5 The IHCI-CAT system is constructed as an example of the main required modules. According to the system disclosed herein, a more advanced approach can be adopted. Figure 5 The system may include more or fewer modules to perform additional operations. Furthermore, a system according to this disclosure may comprise a single module configured to perform two or more operations, or separate modules for each individual operation. Additionally, modules may be implemented in hardware, firmware, software, or any combination thereof.

[0075] The modules and their combinations in the block diagram and / or flowchart can be implemented using computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, and / or other programmable data processing equipment to make it a machine, so that the processor of the computer and / or other programmable data processing equipment can execute these instructions to implement the functions or actions specified in the block diagram and / or flowchart.

[0076] The functions / operations shown in the blocks of a flowchart may occur in a different order than those shown in the operation diagram. For example, two consecutively displayed boxes may actually be executed simultaneously, or the boxes may sometimes be executed in reverse order, depending on the functions / operations involved. Although some diagrams may include arrows on the communication path to indicate the main communication direction, the communication may also proceed in the opposite direction to the arrows shown.

[0077] Furthermore, the solutions disclosed herein can take the form of a computer program on memory, having computer-usable or computer-readable program code implemented in a medium for use by or in conjunction with an instruction execution system. In the context of this document, memory can be any medium containing, storing, or adapted to transfer a program for use by or in conjunction with an instruction execution system, device, or apparatus.

[0078] Therefore, this disclosure also provides a system 600, which includes a processor 601 and a memory 602, such as Figure 6 As shown. In system 600, memory 602 stores instructions that, when executed by processor 601, cause system 600 to perform the method described in the above embodiments.

[0079] This disclosure also provides a computer-readable medium (not shown) storing instructions that, when executed on a system, cause the system to perform the methods described in the above embodiments.

[0080] According to this disclosure, the system can be implemented as a network component on dedicated hardware, as a software instance or firmware running on hardware, as a virtualization function instantiated on a suitable platform (e.g., cloud infrastructure), or in any combination of the above forms.

[0081] Although this specification includes many specific implementation details, these details should not be construed as limiting any implementation or the scope of the claims, but rather as a description of features that may be included in a particular implementation. Certain features described herein as independent embodiments may also be implemented in combination in a single embodiment. Conversely, various features described in a single embodiment may also be implemented individually or in any suitable sub-combination in multiple embodiments. Furthermore, although certain features function in a particular combination in the foregoing description, and even initially claimed as such a combination, in certain circumstances one or more features in a claim may be excluded from that combination, and the claimed combination may also refer to that sub-combination or a variation thereof.

[0082] With the advancement of technology, the inventive concept can be implemented in various ways. The above embodiments are for description only and not for limiting the present disclosure, and modifications and variations can be made without departing from the spirit and scope of the present disclosure. Such modifications and variations are all considered to fall within the scope of the present disclosure and its appended claims. The scope of protection of the present disclosure is defined by the appended claims.

Claims

1. A method for adaptively assessing users' career interest levels using self-comparison questions, comprising the following steps: Establish a question bank containing hundreds of high-quality MFC questions and calibrate the question parameters; each question consists of two statements that are similar in social expectancy, involving two different occupational types, and has K options reflecting K preference categories between the two statements; Initial values ​​for each user's career interests are set before the test is administered. Select new questions from the question bank that are suitable for each user's specific level of career interest, while controlling the exposure of the questions; Based on the user's answers to the selected questions, the estimate of their career interest level is updated, where the user's answer specifically refers to one of the K options; Determine whether the number of test questions administered has reached the preset test length; and If the termination criteria are met, the career interest level is output as the result of the assessment; otherwise, steps 3 and 4 are repeated.

2. The method of claim 1, further comprising the following steps: We constructed a test bank and used an item response theory (IRT) model to calibrate the item parameters of all items on a representative sample of users to determine the attractiveness of the profession.

3. The method of claim 1, wherein the initial value of each user's occupational interest level is set to 0 before the test.

4. The method of claim 1, wherein the selection of the subject includes: New questions are selected based on a function that combines random and statistical information components; moreover, as the test progresses, the influence of the random component on question selection gradually decreases, while the influence of the information component gradually increases.

5. The method of claim 4, wherein the topic selection is performed by function f v To execute, the function is defined as: in, l is the number of items administered, T is the pre-set test length, v contains a vector of identifiers of items in the item bank that were not selected during the current administration, and det(I v R is the determinant of the Fisher Information (FI) matrix for unselected questions. v It is a vector of random numbers generated for each unselected topic from a uniform distribution [0, max{det(Iv)}], where the determinant of the topic FI matrix is ​​calculated using the updated career interest level and topic parameters.

6. The method of claims 4 and 5, further comprising: If a statement is shown to a user the maximum number of times it has been shown, all questions related to that statement will be removed from the question bank.

7. The method of claim 1, wherein the update comprises updating the level of career interest using a Newton-Raphson procedure as follows: Where θ is the vector for estimating career interests; It is a provisional θ estimate after administering l questions. It is an updated θ estimate; furthermore, and These are the first and second derivatives of the natural logarithm of the posterior density function of the reaction vector x, and their calculations are as follows: as well as Where P k v is the probability of choosing category k (k = 0, 1, 2, ..., K-1). k is the total score (sum of scores) for category k, and μ and Φ are the mean vector and variance-covariance matrix of the multivariate normal distribution of θ, respectively.

8. The method of claim 1, wherein the level of occupational interest refers to the six occupational types in Holland's theory of occupational interests: Realistic (R), Investigative (I), Artistic (A), Social (S), Enterprising (E), and Conventional (C).

9. The method of claim 1, wherein the evaluation is an online computer adaptability test (CAT) evaluation.

10. An online adaptability testing system for self-comparison assessment of users' career interest levels, comprising: processor; as well as The system executes the method described in claims 1-9 by storing instructions in memory, which, when executed by the processor.

Images