A method and system for generating MBTI psychometric scale for specific scenarios

Abstract

The application relates to the technical field of psychological assessment, and discloses a method and system for generating an MBTI psychological scale in the face of a specific scene, which comprises the following steps: constructing a scene knowledge network, and constructing a question bank containing multiple items; analyzing an input specific scene description, querying the scene knowledge graph, and generating a multidimensional importance weight matrix for the specific scene through a large language model; screening a candidate question set from the question bank based on the multidimensional importance weight; constructing a comprehensive optimization score function and solving the function through an optimization algorithm to select an optimal item combination from the candidate question set; and arranging the selected optimal item combination to generate a complete MBTI psychological scale. Through the specific scene knowledge network construction mechanism and the scene-based scale generation, the application realizes dynamic generation and real-time optimization of the MBTI psychological scale in a specific application scene, and improves the scene adaptability and result stability of psychological assessment.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of psychological assessment, and in particular to a method and system for generating an MBTI psychological scale in the face of a specific scene. BACKGROUND

[0002] In recent years, psychological assessment tools have been widely used in education management, career planning and personalized services, etc. MBTI (Myers-Briggs Type Indicator) as a commonly used personality classification tool, describes the cognitive and behavioral patterns of individuals by combining their preferences in four dimensions: extroversion / introversion, sensing / intuition, thinking / emotion, and judgment / perception. However, existing MBTI assessment methods rely heavily on fixed questionnaires or static scene descriptions, and lack the ability to dynamically adapt to specific application scenarios. For example, in the context of career counseling or personalized recommendation, an individual's psychological performance is often influenced by the context, task load and social scene, making it difficult for traditional static scales to effectively capture these situational differences. Existing research has attempted to infer personality traits based on social media data, such as using like records, text expressions and other behavioral characteristics to construct low-dimensional representations, and then using logistic regression or support vector machines for personality prediction. Some methods also use the Big Five personality theory as an intermediate layer to indirectly infer MBTI-related types through language features or behavioral characteristics. However, these methods mostly rely on general models and lack adaptive design for specific application tasks, which can easily produce prediction biases when crossing contexts. Moreover, existing scale generation methods rely on fixed question banks and cannot dynamically generate items based on scene requirements, resulting in a semantic gap between the questionnaire and the target task. In summary, existing solutions still have shortcomings in the combination of personality assessment and application scenarios, mainly in the lack of scene-specific feature constraints and the lack of dynamic response to changes in user psychological state, which limits the precise and dynamic application of MBTI scales in specific scenarios.

[0003] Current methods based on fixed questionnaires and general personality modeling have obvious shortcomings in terms of scene adaptability and semantic expression accuracy. These shortcomings are manifested in the following aspects:

[0004] 1. Lack of scene adaptability, low content validity: Traditional MBTI scales aim for universality, and their items are designed around general scenarios. When applied to a highly specific scenario, there is a significant content gap between the questionnaire and the actual assessment requirements. The composition and distribution of the items cannot reflect the differentiated requirements of different psychological characteristics in specific scenarios, resulting in weak correlation between the assessment results and the actual scene performance, and insufficient content validity.

[0005] 2. The rigidity of the scale combination and the lack of optimization mechanism: The item combination of the existing scale is fixed, and it is impossible to adjust the weight and ratio of each dimension question according to the scene demand. There is a lack of a scientific and quantifiable combination optimization mechanism to ensure that the whole questionnaire is the most efficient set for a specific target.

[0006] 3. High customization cost and low efficiency: To solve the above problems, the only traditional way is to hire a psychological measurement expert to manually customize the scale. This process not only has extremely high cost and long time (usually takes weeks or even months), but also highly depends on the personal experience of the expert, lacks unified and reproducible scientific standards. This makes it almost impossible in practice to provide high-quality customized scales for diverse specific scenes.

[0007] 4. Insufficient discrimination of key traits: In a specific scene, the accurate identification of certain key personality traits is crucial. The general scale is far from being able to conduct in-depth and reliable evaluation of individuals, and lacks a mechanism to actively identify and strengthen the investigation of key traits in the scene. SUMMARY

[0008] In view of the rigidity of the existing MBTI scale in content composition, the disconnection with the application scene, and the inefficiency of the customization process, the present application provides a method and system for generating an MBTI psychological scale for a specific scene, aiming to abandon the fixed questionnaire mode, and automatically and scientifically generate an optimal complete scale from a large and high-quality item bank for any specific scene based on a quantitative correlation model between scene requirements and psychological measurement indicators. Specifically, the present application aims to solve the following two core technical problems:

[0009] 1. Realize the deep coupling of scale content and scene demand, that is, establish a set of psychological demand that can quantitatively analyze the specific application scene, and use it as the core basis to intelligently guide the combination and construction of the scale in terms of question selection, dimension ratio, content emphasis, etc., so as to ensure that the finally generated scale has high content validity and scene pertinence.

[0010] 2. Ensure the scientificity and optimality of the generated scale, that is, establish a multi-objective optimization framework, which can simultaneously consider scene relevance, dimension balance, information value (discrimination), content non-redundancy and other key psychological measurement indicators in the decision-making process of generating the scale, so as to ensure that the finally output is not only relevant, but also scientific, efficient and robust.

[0011] To solve the above technical problems, the present application adopts the following technical solutions:

[0012] In a first aspect, the present application provides a method for generating an MBTI psychological scale for a specific scene, comprising:

[0013] Construct a scene knowledge network that includes a scene knowledge graph and a vectorized question bank. It contains multiple items, each labeled with measurement dimensions, information value, scene relevance, and semantic vector;

[0014] Parse the specific scenario description of the input The system queries the scene knowledge graph and generates a multi-dimensional importance weight matrix for a specific scene using a large language model.

[0015] A set of candidate questions is selected from the question bank based on multi-dimensional importance weights;

[0016] Construct a comprehensive optimization scoring function, which includes at least scenario relevance, dimensional balance, information value, and semantic redundancy, and solve it through an optimization algorithm to select the optimal combination of items from the candidate item set;

[0017] The selected optimal combination of items is arranged to generate a complete MBTI psychological scale.

[0018] In one embodiment, the parsed input is a specific scenario description. It queries the scene knowledge graph and generates a multi-dimensional importance weight matrix for a specific scene using a large language model, specifically including:

[0019] The user inputs a specific scenario description. The scenario description is parsed using a large language model, and a knowledge graph is queried to output an MBTI dimension importance weight matrix. :

[0020] ;

[0021] ;

[0022] in, For dimension Importance weight, This indicates an individual's primary tendency to obtain energy. It indicates an individual's tendency in acquiring information; It indicates an individual's tendency in decision-making; It indicates an individual's lifestyle preferences.

[0023] In one embodiment, the step of selecting a set of candidate questions from the question bank based on multi-dimensional importance weights specifically includes:

[0024] Based on the importance weights of different dimensions, from the question bank Selecting candidate topics The higher the importance weight of the dimension, the more candidate questions corresponding to the dimension, and the higher the quality requirement.

[0025] In one embodiment, the comprehensive optimization score function is constructed, including at least scene relevance, dimension balance, information value and semantic redundancy, and specifically including:

[0026] The optimal question to be selected from the candidate question set is modeled as a combinatorial optimization problem, and a comprehensive optimization score is defined as:

[0027] ;

[0028] Scene relevance ; represents the i-th question item, represents the scene relevance of ;

[0029] Dimension balance ; wherein, is the number of questions in the question bank belonging to the dimension , and is an adjustment coefficient;

[0030] Information value ; represents the information value of ;

[0031] Semantic redundancy ; are semantic vectors of , respectively, represents the j-th question item;

[0032] wherein, is a weight parameter.

[0033] In one embodiment, the optimal question combination is selected from the candidate question set by an optimization algorithm, specifically including:

[0034] The combinatorial optimization problem is solved by an optimization algorithm such as genetic algorithm, beam search or integer programming, to obtain the question combination with the highest total score :

[0035] .

[0036] In a second aspect, the present application provides a system for generating an MBTI psychological scale for a specific scenario, comprising:

[0037] A knowledge network construction module constructs a scenario knowledge network including a scenario knowledge graph and a vectorized question bank Each question item is labeled with a measurement dimension, information value, scenario correlation, and semantic vector.

[0038] A weight matrix generation module analyzes the input specific scenario description , queries the scenario knowledge graph, and generates a multi-dimensional importance weight matrix for the specific scenario using a large language model.

[0039] A question selection module selects a candidate question set from the question bank based on the multi-dimensional importance weight.

[0040] An optimal question item selection module constructs a comprehensive optimization score function including scenario relevance, dimension balance, information value, and semantic redundancy, and solves it using an optimization algorithm to select the optimal question item combination from the candidate question set.

[0041] A scale generation module arranges the selected optimal question item combination to generate a complete MBTI psychological scale.

[0042] In one embodiment, the analysis of the input specific scenario description , querying the scenario knowledge graph, and generating a multi-dimensional importance weight matrix for the specific scenario using a large language model, specifically includes:

[0043] The user inputs a specific scenario description , the scenario description is analyzed by a large language model, and the knowledge graph is queried to output an MBTI dimension importance weight matrix :

[0044] ;

[0045] ;

[0046] wherein is the importance weight of the dimension , represents the individual's main energy acquisition tendency, represents the individual's information acquisition tendency; represents the individual's decision-making tendency; represents the individual's lifestyle tendency.

[0047] In one embodiment, the candidate question set is selected from the question bank based on the multi-dimensional importance weight, specifically including:

[0048] According to the importance weight of different dimensions, the question bank​ Screening candidate question set The higher the importance weight of the dimension, the more candidate questions corresponding to the dimension, and the higher the quality requirement.

[0049] In one embodiment, the comprehensive optimization score function is constructed, including at least scene relevance, dimension balance, information value and semantic redundancy, and specifically including:

[0050] The selection of the optimal question from the candidate question set is modeled as a combinatorial optimization problem, and a comprehensive optimization score is defined As follows:

[0051] ;

[0052] Scene relevance ;The i-th question item is represented as The scene relevance of is represented as

[0053] Dimension balance ; wherein The number of questions in the question bank belonging to the dimension is represented as is an adjustment coefficient;

[0054] Information value ;The information value of is represented as

[0055] Semantic redundancy ;The semantic vector of is represented as The j-th question item is represented as Wherein

[0056] is a weight parameter. In one embodiment, the optimal question combination is selected from the candidate question set by solving the optimization algorithm, specifically including:

[0057] By using optimization algorithms such as genetic algorithm, beam search or integer programming, the combinatorial optimization problem is solved to obtain the question combination with the highest total score

[0058] :

[0059] .

[0060] Compared with the prior art, the beneficial technical effects of the present application are:

[0061] The application realizes the dynamic generation and real-time optimization of the MBTI psychological scale in a specific application scenario through the specific scene knowledge network construction mechanism and the scene-based scale generation method, and significantly improves the situational adaptability and result stability of psychological assessment. Specifically, the following advantages are mainly achieved:

[0062] First, the depth coupling of scale content and application scenario is realized: each generated scale is customized for a specific application scenario, has high content validity, and directly solves the problem of insufficient customization of general scales.

[0063] Second, the psychological measurement quality of the generated scale is ensured: through multi-objective optimization, the system not only considers the scene correlation, but also considers the dimension balance, information value and non-redundancy and other core psychological measurement indicators, ensuring the scientificity and robustness of the generated scale.

[0064] Third, high flexibility and scalability: users can generate exclusive scales for any emerging and specific scenario at any time according to their needs. At the same time, by continuously expanding and optimizing the knowledge graph and question bank in the back end, the system's capabilities can be continuously enhanced. BRIEF DESCRIPTION OF DRAWINGS

[0065] Figure 1 The method flowchart in the embodiment of the application is shown in the figure;

[0066] Figure 2 The overall technical framework schematic diagram in the embodiment of the application is shown in the figure;

[0067] Figure 3 The schematic diagram of the scene-based generation mechanism in the embodiment of the application is shown in the figure. DETAILED DESCRIPTION

[0068] A preferred embodiment of the application will be described in detail below with reference to the accompanying drawings.

[0069] As shown in the figure, Figure 1 a method for generating an MBTI psychological scale for a specific scenario in the application includes the following steps:

[0070] S1, a scene knowledge network containing a scene knowledge graph and a vectorized question bank is constructed, the question bank contains a plurality of items, and each item is labeled with a measurement dimension, an information value, a scene correlation degree and a semantic vector;

[0071] S2, analyze the input specific scene description , and query the scene knowledge graph to generate a multi-dimensional importance weight matrix for the specific scene through a large language model;

[0072] S3, screening a candidate question set from the question bank based on the importance weight of multiple dimensions;

[0073] S4, constructing a comprehensive optimization score function including at least scenario relevance, dimension balance, information value and semantic redundancy, and solving it by optimization algorithm to select the optimal question combination from the candidate question set;

[0074] S5, arranging the selected optimal question combination to generate a complete MBTI psychological scale.

[0075] The overall technical framework of the present application is shown in Figure 2 The core goal of the present method is to automatically and instantly generate a set of robust complete MBTI questionnaire in psychology for any given application scenario. To achieve this goal, the present framework introduces a psychological knowledge graph as the knowledge base, and designs a multi-objective combination optimization function. Through this method, a complete MBTI scale that can be directly used and deeply optimized for a specific scenario is output, thereby solving the fundamental problem of the disconnection between traditional general scale and application requirements.

[0076] The present application is mainly realized through two processes, including specific scenario knowledge network construction and scenario-based scale generation.

[0077] 1. Specific scenario knowledge network construction.

[0078] First, a knowledge network of "scene-dimension-question" association is constructed, including a scenario knowledge graph and a vectorized question bank. The knowledge graph defines the highly associated dimensions of a specific scenario. Then a vectorized question bank is established, creating a massive question bank containing tens of thousands of standard MBTI questions, and each question is annotated and vectorized with rich metadata. .

[0079] Each question has the following attributes:

[0080] (1) Dimension: ; (Extraversion–Introversion, extraversion–introversion) represents the individual's tendency to obtain energy, extraversion refers to the preference for obtaining energy from external activities and interpersonal interactions, introversion refers to the preference for obtaining energy from internal thinking and solitude; (Sensing–iNtuition, Sensing–iNtuition) represents the individual's tendency in information acquisition, Sensing refers to the preference for perceiving the world through concrete facts and details, intuition refers to the preference for understanding the world through abstract patterns and overall relationships; (Thinking–Feeling) represents an individual's tendency in decision-making; thinking... Preferences refer to decisions based on logic and objective criteria, rather than emotions. This refers to a preference for making decisions based on values ​​and interpersonal emotions. (Judging–Perceiving) represents an individual's lifestyle preferences, specifically judgment. This refers to a preference for a planned and structured lifestyle, and perception. It refers to a preference for a flexible, open, and adaptable lifestyle.

[0081] (2) Information value (discrimination): ;

[0082] (3) Scene relevance: ;

[0083] (4) Semantic vector: .

[0084] Questions semantic similarity between .

[0085] 2. Contextualized scale generation.

[0086] Contextualized generation mechanism, such as Figure 3 As shown.

[0087] The user inputs a specific scenario description. The scene description is parsed using a large language model, and a knowledge graph is queried to output an MBTI dimension importance weight matrix:

[0088] ;

[0089] .

[0090] in Dimensions in the Scene Importance weight, Then, based on the importance weight of each dimension, in Selecting candidate topics Dimensions with higher importance weights have more candidate questions and higher quality requirements. The selection of the optimal question from the candidate question set is modeled as a combinatorial optimization problem, and a comprehensive optimization score is defined. for:

[0091] .

[0092] The definitions of each part are as follows:

[0093] (1) Scenario Relevance .

[0094] (2) Dimension Balance . Wherein is the number of questions in the topic set that belong to the dimension , is the adjustment coefficient.

[0095] (3) Information Value .

[0096] (4) Semantic Redundancy .

[0097] Wherein, is the weight parameter, used to balance the importance of each optimization objective.

[0098] Then, through genetic algorithm, beam search or integer programming optimization algorithm, the combined optimization problem is solved, and finally a total score highest question item combination is obtained:

[0099] .

[0100] Finally, the question item combination selected by optimization is randomly sorted or arranged according to a specific logic, and is packaged into a complete MBTI scale that can be directly answered by the end user (for example, a PDF file or a web page link is generated).

[0101] The present application systematically solves the problems of insufficient scenario adaptability, limited semantic alignment capability and result update lag through the following key mechanisms:

[0102] (1) The knowledge of psychometrics, scenario requirements and item bank resource structure are structured into knowledge graph, and the systematic method of scenario analysis, dimension weighting and candidate question selection is based on this. It provides high-quality input and constraints for the generation of scale by combination optimization, and is the basis for realizing high-precision scenario customization.

[0103] (2) Through combination optimization generation, instead of directly generating text, a multi-objective optimization function is used to comprehensively consider scenario relevance, dimension balance, information value, semantic redundancy, etc. to select and combine an optimal question item set from a large-scale question bank, thereby forming a complete scale. This is the key innovation that distinguishes it from traditional fixed scales and simple rule screening.

[0104] (3) The application integrates the above mechanism into an end-to-end dynamic scale generation method, which can effectively cope with semantic differences, uneven distribution of psychological characteristics and dynamic evolution of user states in different application scenarios, and significantly improve the adaptability, stability and interpretability of psychological assessment.

[0105] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. As used herein, the terms "comprises", "comprising", "includes", "including" and the like are specifically intended to be open-ended and to mean that other features, steps, operations, and / or components can be added.

[0106] It should be understood that, although the steps in the flowcharts of the drawings are shown in a sequential order, these steps are not necessarily performed in the order shown by the arrows. Unless otherwise specified herein, the steps are not necessarily performed in the order shown, and these steps can be performed in other orders. Moreover, at least some of the steps in the flowcharts of the drawings can include multiple steps or multiple stages, which are not necessarily performed at the same time, but can be performed at different times, and the order of the steps or stages is not necessarily sequential, but can be performed in rotation or alternation with at least some of the steps or stages in other steps or stages.

[0107] Based on the description of the method embodiments, the application further provides a system. The system can be a system using the method described in the embodiments of the present application, such as software (application), module, component, server, client, etc., in combination with necessary implementation hardware. Based on the same innovative concept, the system in one or more embodiments provided by the embodiments of the present application is described as follows. Since the implementation scheme of the system to solve the problem is similar to the method, the implementation of the specific system in the embodiments of the present application can be referred to the implementation of the foregoing method, and the repeated parts will not be described herein. The term "module" or "module" used below is a combination of software and / or hardware that can realize a predetermined function. Although the system described in the following embodiments is preferably realized in software, the realization of hardware or a combination of software and hardware is also possible and is conceived.

[0108] The technical features of the above embodiments can be combined in any manner. In order to make the description concise, not all possible combinations of the technical features in the above embodiments are described, however, as long as the combination of the technical features does not exist contradictory, it should be considered as the scope of the present application.

[0109] It will be obvious to a person skilled in the art that the application is not limited to the details of the foregoing exemplary embodiments and can be implemented in other concrete forms without departing from the spirit or essential characteristics of the application. The embodiments are therefore to be considered in all respects as illustrative and not restrictive, the scope of the application being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein and no

[0110] Furthermore, it should be understood that although the description is made on the basis of embodiments, not every embodiment contains only one independent technical solution, and the description is made in this way only for the sake of clarity, and a person skilled in the art should consider the description as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by a person skilled in the art.

Claims

1. A method for generating an MBTI psychological scale for a specific scenario, characterized in that, include: Construct a scene knowledge network that includes a scene knowledge graph and a vectorized question bank. It contains multiple items, each labeled with measurement dimensions, information value, scene relevance, and semantic vector; Parse the specific scenario description of the input The system queries the scene knowledge graph and generates a multi-dimensional importance weight matrix for a specific scene using a large language model. A set of candidate questions is selected from the question bank based on multi-dimensional importance weights; Construct a comprehensive optimization scoring function, which includes at least scenario relevance, dimensional balance, information value, and semantic redundancy, and solve it through an optimization algorithm to select the optimal combination of items from the candidate item set; The selected optimal combination of items is arranged to generate a complete MBTI psychological scale.

2. The method for generating an MBTI psychological scale for a specific scenario according to claim 1, characterized in that, The specific scenario description of the parsed input It queries the scene knowledge graph and generates a multi-dimensional importance weight matrix for a specific scene using a large language model, specifically including: The user inputs a specific scenario description. The scenario description is parsed using a large language model, and a knowledge graph is queried to output an MBTI dimension importance weight matrix. : ； ； in, For dimension Importance weight, This indicates an individual's primary tendency to obtain energy. It indicates an individual's tendency in acquiring information; It indicates an individual's tendency in decision-making; It indicates an individual's lifestyle preferences.

3. The method for generating an MBTI psychological scale for a specific scenario according to claim 1, characterized in that, The selection of candidate question sets from the question bank based on multi-dimensional importance weights specifically includes: Based on the importance weights of different dimensions, from the question bank Selecting candidate topics The higher the importance weight of a dimension, the more candidate topics there are, and the higher the quality requirements.

4. The method for generating an MBTI psychological scale for a specific scenario according to claim 2, characterized in that, The construction of a comprehensive optimization scoring function includes at least scene relevance, dimensional balance, information value, and semantic redundancy, specifically including: This will require selecting from the candidate topic set. The optimal problem is selected and modeled as a combinatorial optimization problem, and a comprehensive optimization score is defined. for: ； Scene relevance ; This represents the i-th item. express The degree of relevance to the scene; Dimensional balance ;in, For the question bank, it belongs to the dimension The number of questions This is the adjustment coefficient; Information value ; express Information value; semantic redundancy ; They are respectively semantic vectors, This represents the j-th item; in, These are the weight parameters.

5. The method for generating an MBTI psychological scale for a specific scenario according to claim 4, characterized in that, The step of selecting the optimal combination of questions from the candidate question set through optimization algorithms specifically includes: Solve the combinatorial optimization problem using genetic algorithms, beam search algorithms, or integer programming algorithms to obtain the combination of questions with the highest total score. : 。 6. A system for generating an MBTI psychological scale for a specific scenario, characterized in that, include: The knowledge network construction module builds a scene knowledge network that includes a scene knowledge graph and a vectorized question bank. It contains multiple items, each labeled with measurement dimensions, information value, scene relevance, and semantic vector; The weight matrix generation module parses the input specific scene description. The system queries the scene knowledge graph and generates a multi-dimensional importance weight matrix for a specific scene using a large language model. The question selection module filters a set of candidate questions from the question bank based on multi-dimensional importance weights; The optimal question selection module constructs a comprehensive optimization scoring function, which includes at least scenario relevance, dimensional balance, information value, and semantic redundancy, and solves it through an optimization algorithm to select the optimal combination of questions from the candidate question set. The scale generation module arranges the selected optimal combination of items to generate a complete MBTI psychological scale.

7. The system for generating an MBTI psychological scale for a specific scenario according to claim 6, characterized in that, The specific scenario description of the parsed input It queries the scene knowledge graph and generates a multi-dimensional importance weight matrix for a specific scene using a large language model, specifically including: The user inputs a specific scenario description. The scenario description is parsed using a large language model, and a knowledge graph is queried to output an MBTI dimension importance weight matrix. : ； ； in, For dimension Importance weight, This indicates an individual's primary tendency to obtain energy. It indicates an individual's tendency in acquiring information; It indicates an individual's tendency in decision-making; It indicates an individual's lifestyle preferences.

8. The system for generating an MBTI psychological scale for a specific scenario according to claim 6, characterized in that, The selection of candidate question sets from the question bank based on multi-dimensional importance weights specifically includes: Based on the importance weights of different dimensions, from the question bank Selecting candidate topics The higher the importance weight of a dimension, the more candidate topics there are, and the higher the quality requirements.

9. A system for generating an MBTI psychological scale for a specific scenario according to claim 7, characterized in that, The construction of a comprehensive optimization scoring function includes at least scene relevance, dimensional balance, information value, and semantic redundancy, specifically including: The problem of selecting the optimal question from the candidate question set is modeled as a combinatorial optimization problem, and a comprehensive optimization score is defined. for: ； Scene relevance ; This represents the i-th item. express The degree of relevance to the scene; Dimensional balance ;in, For the question bank, it belongs to the dimension The number of questions This is the adjustment coefficient; Information value ; express Information value; semantic redundancy ; They are respectively semantic vectors, This represents the j-th item; in, These are the weight parameters.

10. A system for generating an MBTI psychological scale for a specific scenario according to claim 9, characterized in that, The step of selecting the optimal combination of questions from the candidate question set through optimization algorithms specifically includes: Solve the combinatorial optimization problem using genetic algorithms, beam search algorithms, or integer programming algorithms to obtain the combination of questions with the highest total score. : 。

Images