A question generation method and device, computer equipment and storage medium

Abstract

The present disclosure provides a question generation method and device, computer equipment and a storage medium, wherein the method comprises: in response to a word exercise instruction, determining a question type proportion corresponding to each question type level of a plurality of preset question type levels; different question type levels are used to test different word abilities; determining a correct answer rate of a target user for each candidate word based on an ability coefficient of the target user under the question type proportion; the ability coefficient of the target user under the question type proportion is determined based on historical answer results of the target user for target historical questions of the question type level corresponding to the question type proportion; based on the number of words carried by the word exercise instruction, the correct answer rate and the mastering question type level of the target user for the candidate word, a target word meeting the question type proportion is selected from a plurality of candidate words; generating a target question corresponding to the target word, and displaying the target question through a target user terminal.

Description

Technical Field

[0001] This disclosure relates to the field of computer technology, and more specifically, to a question generation method, apparatus, computer device, and storage medium. Background Technology

[0002] Currently, more and more people are choosing to use vocabulary learning software. Typically, this software randomly selects words from a user-selected pre-defined vocabulary list and then searches for corresponding questions for the user to answer. However, because the words are randomly chosen from the pre-defined list, the selected words can easily be too easy or too difficult, resulting in questions that are either too easy or too difficult, thus reducing the user's learning experience and efficiency. Therefore, generating questions of appropriate difficulty has become a pressing issue. Summary of the Invention

[0003] This disclosure provides at least one method, apparatus, computer device, and storage medium for generating questions.

[0004] In a first aspect, embodiments of this disclosure provide a question generation method, including:

[0005] In response to the vocabulary practice instruction, the system determines the proportion of each of the multiple preset question levels; different question levels are used to assess different vocabulary abilities.

[0006] Based on the target user's ability coefficient under the specified question category ratio, the target user's correct answer rate for each candidate word is determined; wherein, the target user's ability coefficient under the specified question category ratio is determined based on the target user's historical answer results for target question categories at the target question category level corresponding to the specified question category ratio;

[0007] Based on the number of words practiced in the word practice instruction, the correct answer rate, and the target user's mastery level of the candidate words by question type, target words that meet the question type ratio are selected from multiple candidate words;

[0008] Generate the target question corresponding to the target word and display the target question through the target user terminal.

[0009] In one possible implementation, determining the proportion of question types corresponding to multiple preset question type levels includes:

[0010] Based on the mastery level of each candidate word, the proportion of each question type corresponding to a number of preset question types is determined.

[0011] In one possible implementation, after determining the proportion of question types corresponding to multiple preset question type levels, the method further includes:

[0012] If the question category ratio includes a first question category level and other question category levels besides the first question category level, and the first question category ratio corresponding to the first question category level exceeds a preset ratio, then the first question category ratio corresponding to the first question category level and the second question category ratio corresponding to the other question category levels are adjusted.

[0013] In one possible implementation, the method further includes determining the target user's ability coefficient under the proportion of the question type according to the following method:

[0014] Based on the historical answers to the target historical questions and the word difficulty of the candidate words contained in the target historical questions, the ability coefficient of the target user under the proportion of the question type is determined.

[0015] In one possible implementation, determining the target user's correct answer rate for each candidate word based on the target user's ability coefficient under the proportion of the question type includes:

[0016] Based on the target user's ability coefficient under the question category ratio, the word difficulty of each candidate word, and the preset target parameters, the target user's correct answer rate for each candidate word is determined;

[0017] The target parameter represents the degree of influence of the historical browsing status of each candidate word on the correct answer rate. The target parameter is determined based on the statistical results of the answers of multiple first test users to browsed words and the statistical results of the answers to unbrowsed words.

[0018] In one possible implementation, the method further includes determining the word difficulty of each of the candidate words according to the following method:

[0019] Based on the statistical results of multiple second test users' answers to the questions corresponding to each candidate word, the word difficulty of each candidate word is determined.

[0020] In one possible implementation, if the number of first statistical users exceeds a preset number, the plurality of second test users are the first statistical users; if the number of first statistical users does not exceed the preset number, the plurality of second test users are the second statistical users; or they are the first statistical users and the second statistical users, wherein the attribute information of the first statistical users is the same as that of the target user, and the attribute information of the second statistical users is different from that of the target user.

[0021] In one possible implementation, the candidate words are selected from the multimedia content that the target user has browsed in the past;

[0022] The step of selecting target words that meet the question category ratio from multiple candidate words based on the number of words practiced in the word practice instruction, the correct answer rate, and the target user's mastery level of the candidate words includes:

[0023] Based on the number of words practiced in the word practice instruction, the correct answer rate, the target user's mastery level of the candidate words, and the target time corresponding to each candidate word, target words that meet the question type ratio are selected from multiple candidate words; wherein, the target time corresponding to each candidate word includes the selected time corresponding to each candidate word and / or the historical practice time corresponding to each candidate word.

[0024] In one possible implementation, generating the target question corresponding to the target word includes:

[0025] Identify at least one candidate question type corresponding to the mastery level of the target word;

[0026] Based on the priority of the at least one candidate question type, the target question type corresponding to the target word is determined from the at least one candidate question type;

[0027] Based on the target words and the target question type, the target question is generated.

[0028] In one possible implementation, generating the target question based on the target word and the target question type includes:

[0029] When the target question type is multiple choice, determine the target distractor word corresponding to the target word;

[0030] Based on the target distractor word, the target question type, and the target word, generate the target question corresponding to the target word.

[0031] In one possible implementation, determining the target interference word corresponding to the target word includes:

[0032] Determine the preset candidate distractor words corresponding to the target word, as well as the historical words contained in the target user's historical questions; wherein, the historical words include historical practice words and / or historical distractor words;

[0033] From the candidate interference words, select the target interference words other than the historical words.

[0034] In one possible implementation, after displaying the target question to the target user, the method further includes:

[0035] Obtain the target user's target answer to the target question;

[0036] Based on the target answer results and the preset question category level update rules, the mastery question category level of the target words is updated; wherein, the question category level update rules are used to indicate the method for changing the mastery question category level of each target word when the corresponding target question is answered correctly or incorrectly.

[0037] Secondly, this disclosure also provides a question generation apparatus, comprising:

[0038] The first determining module is used to respond to the word practice instruction and determine the proportion of question types corresponding to multiple preset question levels; among them, different question levels are used to test different word abilities;

[0039] The second determining module is used to determine the target user's correct answer rate for each candidate word based on the target user's ability coefficient under the question category ratio; wherein, the target user's ability coefficient under the question category ratio is determined based on the target user's historical answer results for target question category level corresponding to the question category ratio;

[0040] The filtering module is used to filter target words that meet the question category ratio from multiple candidate words based on the number of words practiced in the word practice instruction, the correct answer rate, and the target user's mastery level of the candidate words.

[0041] The generation module is used to generate target questions corresponding to the target words and display the target questions through the target user terminal.

[0042] In one possible implementation, the first determining module, when determining the proportion of question types corresponding to multiple preset question type levels, is used to:

[0043] Based on the mastery level of each candidate word, the proportion of each question type corresponding to a number of preset question types is determined.

[0044] In one possible implementation, after determining the proportion of question types corresponding to multiple preset question type levels, the first determining module is further configured to:

[0045] If the question category ratio includes a first question category level and other question category levels besides the first question category level, and the first question category ratio corresponding to the first question category level exceeds a preset ratio, then the first question category ratio corresponding to the first question category level and the second question category ratio corresponding to the other question category levels are adjusted.

[0046] In one possible implementation, the device is also used to determine the target user's ability coefficient under the proportion of the question type according to the following method:

[0047] Based on the historical answers to the target historical questions and the word difficulty of the candidate words contained in the target historical questions, the ability coefficient of the target user under the proportion of the question type is determined.

[0048] In one possible implementation, the second determining module, when determining the target user's correct answer rate for each candidate word based on the target user's ability coefficient under the proportion of the question type, is used to:

[0049] Based on the target user's ability coefficient under the question category ratio, the word difficulty of each candidate word, and the preset target parameters, the target user's correct answer rate for each candidate word is determined;

[0050] The target parameter represents the degree of influence of the historical browsing status of each candidate word on the correct answer rate. The target parameter is determined based on the statistical results of the answers of multiple first test users to browsed words and the statistical results of the answers to unbrowsed words.

[0051] In one possible implementation, the device is also used to determine the word difficulty of each of the candidate words according to the following method:

[0052] Based on the statistical results of multiple second test users' answers to the questions corresponding to each candidate word, the word difficulty of each candidate word is determined.

[0053] In one possible implementation, if the number of first statistical users exceeds a preset number, the plurality of second test users are the first statistical users; if the number of first statistical users does not exceed the preset number, the plurality of second test users are the second statistical users; or they are the first statistical users and the second statistical users, wherein the attribute information of the first statistical users is the same as that of the target user, and the attribute information of the second statistical users is different from that of the target user.

[0054] In one possible implementation, the candidate words are selected from the multimedia content that the target user has browsed in the past;

[0055] The filtering module, when filtering target words that meet the question category ratio from multiple candidate words based on the number of words practiced in the word practice instruction, the correct answer rate, and the target user's mastery level of the candidate words, is used for:

[0056] Based on the number of words practiced in the word practice instruction, the correct answer rate, the target user's mastery level of the candidate words, and the target time corresponding to each candidate word, target words that meet the question type ratio are selected from multiple candidate words; wherein, the target time corresponding to each candidate word includes the selected time corresponding to each candidate word and / or the historical practice time corresponding to each candidate word.

[0057] In one possible implementation, the generation module, when generating the target question corresponding to the target word, is used to:

[0058] Identify at least one candidate question type corresponding to the mastery level of the target word;

[0059] Based on the priority of the at least one candidate question type, the target question type corresponding to the target word is determined from the at least one candidate question type;

[0060] Based on the target words and the target question type, the target question is generated.

[0061] In one possible implementation, the generation module, when generating the target question based on the target word and the target question type, is used to:

[0062] When the target question type is multiple choice, determine the target distractor word corresponding to the target word;

[0063] Based on the target distractor word, the target question type, and the target word, generate the target question corresponding to the target word.

[0064] In one possible implementation, the generation module, when determining the target interference word corresponding to the target word, is used to:

[0065] Determine the preset candidate distractor words corresponding to the target word, as well as the historical words contained in the target user's historical questions; wherein, the historical words include historical practice words and / or historical distractor words;

[0066] From the candidate interference words, select the target interference words other than the historical words.

[0067] In one possible implementation, after displaying the target question to the target user, the device is further configured to:

[0068] Obtain the target user's target answer to the target question;

[0069] Based on the target answer results and the preset question category level update rules, the mastery question category level of the target words is updated; wherein, the question category level update rules are used to indicate the method for changing the mastery question category level of each target word when the corresponding target question is answered correctly or incorrectly.

[0070] Thirdly, embodiments of this disclosure also provide a computer device, including: a processor, a memory, and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the computer device is running, the processor communicates with the memory via the bus, and when the machine-readable instructions are executed by the processor, the steps of the first aspect above, or any possible implementation of the first aspect, are performed.

[0071] Fourthly, embodiments of this disclosure also provide a computer-readable storage medium storing a computer program that, when executed by a processor, performs the steps of the first aspect or any possible implementation of the first aspect.

[0072] The question generation method, apparatus, computer device, and storage medium provided in this disclosure can first determine the question category ratios corresponding to multiple preset question category levels after responding to a word practice instruction. Since different question category levels are used to test different word abilities, the target user's correct answer rate for each candidate word can be accurately determined based on the target user's ability coefficient under the question category ratios. Then, based on the number of words practiced in the word practice instruction, the correct answer rate, and the target user's mastery level of the candidate words, target words that meet the question category ratios are selected from the multiple candidate words. In this way, since the difficulty of the target words is suitable for the target user's current level, the target questions generated based on the target words are also of moderate difficulty for the target user, thereby improving the user's learning efficiency and learning experience.

[0073] To make the above-mentioned objects, features and advantages of this disclosure more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0074] To more clearly illustrate the technical solutions of the embodiments of this disclosure, the accompanying drawings used in the embodiments will be briefly described below. These drawings are incorporated in and constitute a part of this specification. They illustrate embodiments conforming to this disclosure and, together with the specification, serve to explain the technical solutions of this disclosure. It should be understood that the following drawings only show some embodiments of this disclosure and should not be considered as limiting the scope. Those skilled in the art can obtain other related drawings based on these drawings without creative effort.

[0075] Figure 1 A flowchart of a question generation method provided by an embodiment of this disclosure is shown;

[0076] Figure 2 A schematic diagram illustrating the overall flow of a question generation method provided in an embodiment of this disclosure is shown.

[0077] Figure 3 This diagram illustrates the architecture of a question generation apparatus provided in an embodiment of the present disclosure.

[0078] Figure 4 A schematic diagram of the structure of a computer device provided in an embodiment of this disclosure is shown. Detailed Implementation

[0079] To make the objectives, technical solutions, and advantages of the embodiments of this disclosure clearer, the technical solutions of the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this disclosure, and not all of them. The components of the embodiments of this disclosure described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this disclosure provided in the accompanying drawings is not intended to limit the scope of the claimed disclosure, but merely represents selected embodiments of this disclosure. All other embodiments obtained by those skilled in the art based on the embodiments of this disclosure without inventive effort are within the scope of protection of this disclosure.

[0080] In related technologies, vocabulary learning software typically allows users to first select a vocabulary database to learn, such as a CET-4 vocabulary database, a CET-6 vocabulary database, or a vocabulary database for a specific exam syllabus. Then, words are randomly selected from the user-selected database to generate questions for the user to answer.

[0081] However, since the words used in the assessment are randomly selected from the vocabulary database, and each user's vocabulary level is different, the randomly selected words may be too easy or too difficult for any given user. This can lead to questions generated based on those words being either too easy or too difficult. If the questions are too easy, it will waste the user's time and reduce their learning efficiency. If the questions are too difficult, it will dampen the user's interest in learning.

[0082] Based on the above research, this disclosure provides a question generation method, apparatus, computer device, and storage medium. First, in response to a word practice instruction, it determines the proportion of question categories corresponding to multiple preset question category levels. Since different question category levels are used to assess different word abilities, the target user's correct answer rate for each candidate word can be accurately determined based on the target user's ability coefficient under the specified question category proportions. Then, based on the number of words practiced in the word practice instruction, the correct answer rate, and the target user's mastery level of the candidate words, target words that meet the specified question category proportions are selected from the multiple candidate words. In this way, because the difficulty of the target words is suitable for the target user's current level, the target questions generated based on the target words are also of moderate difficulty for the target user, thereby improving the user's learning efficiency and learning experience.

[0083] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0084] In this document, the term "and / or" merely describes a relationship, indicating that three relationships can exist. For example, A and / or B can represent three cases: A alone, A and B simultaneously, and B alone. Furthermore, the term "at least one" in this document means any combination of at least two of any one or more elements. For example, including at least one of A, B, and C can mean including any one or more elements selected from the set consisting of A, B, and C.

[0085] It is understood that before using the technical solutions disclosed in the various embodiments of this disclosure, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in this disclosure in an appropriate manner in accordance with relevant laws and regulations, and user authorization should be obtained.

[0086] For example, upon receiving a user's active request, a prompt message is sent to the user to explicitly inform them that the requested operation will require the acquisition and use of the user's personal information. This allows the user to independently choose whether to provide personal information to the software or hardware, such as the electronic device, application, server, or storage medium performing the operations of this disclosed technical solution, based on the prompt message.

[0087] As an optional but non-limiting implementation, in response to a user's active request, sending a prompt message to the user can be done via a pop-up window, where the prompt message can be presented in text format. Furthermore, the pop-up window can also include a selection control allowing the user to choose "agree" or "disagree" to provide personal information to the electronic device.

[0088] It is understood that the above notification and user authorization process are merely illustrative and do not constitute a limitation on the implementation of this disclosure. Other methods that comply with relevant laws and regulations may also be applied to the implementation of this disclosure.

[0089] To facilitate understanding of this embodiment, a question generation method disclosed in this disclosure will first be described in detail. The execution entity of the question generation method provided in this disclosure is generally a user terminal or a server. The user terminal can be, for example, a smartphone, tablet computer, or personal computer. In one possible application scenario, the question generation method provided in this disclosure can be applied to a target application on the user terminal. In some possible implementations, the question generation method can be implemented by a processor calling computer-readable instructions stored in memory.

[0090] See Figure 1 The diagram shows a flowchart of a question generation method provided in this embodiment of the present disclosure. The method includes steps 101 to 104, wherein:

[0091] Step 101: In response to the vocabulary practice instruction, determine the proportion of each of the multiple preset question levels; where different question levels are used to test different vocabulary abilities.

[0092] Step 102: Based on the target user's ability coefficient under the question category ratio, determine the target user's correct answer rate for each candidate word; wherein, the target user's ability coefficient under the question category ratio is determined based on the target user's historical answer results for target question category level corresponding to the question category ratio;

[0093] Step 103: Based on the number of words practiced in the word practice instruction, the correct answer rate, and the target user's mastery level of the candidate words, select target words that meet the question category ratio from multiple candidate words;

[0094] Step 104: Generate the target question corresponding to the target word and display the target question through the target user terminal.

[0095] The following is a detailed explanation of the steps described above:

[0096] Regarding step 101,

[0097] Specifically, the vocabulary practice instruction can be generated in response to a target user's targeted action. This targeted action includes, but is not limited to, single-click, double-click, long-press, swipe, and drag operations. For example, the vocabulary practice instruction can be generated after the user clicks a target button (such as the "Start Practice" button). Alternatively, the vocabulary practice instruction can also be sent from other user terminals or servers, such as from a teacher's terminal to the executing entity. This embodiment does not limit the method for generating the vocabulary practice instruction.

[0098] The vocabulary ability is used to characterize a user's understanding and application of words. For example, the vocabulary ability can be divided into three levels from easy to difficult: recognition, limited application, and free application. Specifically, it can be divided into: sound-form association (i.e., being able to determine the English word based on the pronunciation of the word, and being able to determine the pronunciation of the word based on the English word), sound-meaning association (i.e., being able to determine the Chinese word based on the pronunciation of the word, and being able to determine the pronunciation of the word based on the Chinese word), meaning-form association (i.e., being able to determine the English word based on the Chinese word, and being able to determine the Chinese word based on the English word), form-sound association (i.e., being able to pronounce the word based on the English word), meaning-sound association (i.e., being able to pronounce the word based on the Chinese word), and being able to complete the sound-meaning association and read the word in various contexts.

[0099] The multiple preset question level settings can be based on multiple vocabulary abilities. Since vocabulary abilities vary in difficulty, the question level settings based on vocabulary abilities also correspond to different question difficulty levels. The correspondence between question level, question difficulty, and vocabulary ability can be shown in Table 1 below:

[0100]

[0101] Table 1

[0102] The question category ratio represents the ratio of the number of target words to be practiced corresponding to multiple preset question category levels. For example, the question category ratio is: Question Category Level 1: Question Category Level 2: Question Category Level 3: Question Category Level 4: Question Category Level 5 = 1:2:3:4:5.

[0103] In one possible implementation, when determining the proportion of question categories corresponding to multiple preset question category levels, the proportion of question categories corresponding to multiple preset question category levels can be determined first based on the mastery level of each candidate word.

[0104] Specifically, the "mastery level" represents the level of a question category that the user has reached among the multiple preset question category levels. Each candidate word can correspond to its own mastery level. Then, the ratio of the number of candidate words corresponding to each mastery level can be counted, and the ratio of the number of candidate words can be used as the question category ratio corresponding to each of the multiple preset question category levels.

[0105] For example, if the candidate words include 100 candidate words corresponding to question category level 1, 80 candidate words corresponding to question category level 2, and 20 candidate words corresponding to question category level 5, then the question category ratio is: Question category level 1: Question category level 2: Question category level 3: Question category level 4: Question category level 5 = 5:4:0:0:1.

[0106] Using this method, since the question category ratio is used to control the number of target words at each question category level selected, and the question category ratio is determined by the mastery level of each candidate word, the amount of target word practice at each question category level can be reasonably arranged. Thus, the training amount of each word ability of the target user can be personalized according to the target user's mastery of each candidate word, thereby improving learning efficiency.

[0107] It is understandable that if there are too many questions in a certain question category level, it will not be conducive to the comprehensive improvement of the target user's multiple vocabulary abilities. If there are too many questions in a more difficult question category level, the questions may be too difficult for the target user. Therefore, in one possible implementation, if the question category ratio includes a first question category level and other question category levels besides the first question category level, and the first question category ratio corresponding to the first question category level exceeds a preset ratio, the first question category ratio corresponding to the first question category level and the second question category ratio corresponding to the other question category levels are adjusted.

[0108] Wherein, the first question category ratio represents the ratio of the number of target words to be practiced to the number of word practice sessions corresponding to the first question category level, the second question category ratio represents the ratio of the number of target words to be practiced to the number of word practice sessions corresponding to the other question category levels, and the number of word practice sessions is used to represent the total number of target words to be practiced. The first question category level can be, for example, a question category level with higher difficulty, such as question category level 5 in Table 1.

[0109] Specifically, at least one first question category level can be preset among the multiple preset question category levels, and a preset proportion corresponding to each of the at least one first question category level, such as the first question category level being question category level 5 and the preset proportion being 80%. Then, it is determined whether the question category proportions corresponding to the multiple preset question category levels include the first question category level (or whether the proportion item corresponding to the first question category level in the question category proportion is 0) and other question category levels (or whether the proportion item corresponding to the other question category levels in the question category proportion is 0). If both are included (or if neither is 0), it is determined whether the first question category proportion corresponding to the first question category level exceeds the preset proportion corresponding to the first question category level. If so, the first question category proportion is reduced, and the reduction ratio of the first question category proportion is increased to the second question category proportion.

[0110] Here, when reducing the proportion of the first question category, for example, the proportion of the first question category can be multiplied by a preset weighting coefficient (such as 0.9) to obtain the adjusted proportion of the first question category, or the proportion of the first question category can be reduced to the preset proportion. Then, when increasing the proportion of the second question category, the difference between the adjusted proportion of the first question category and the proportion of the first question category before adjustment (i.e., the reduction proportion) can be added to the proportion of the second question category to obtain the adjusted proportion of the second question category. When there are multiple other question category levels, for example, the reduction proportion can be increased on average to each proportion of the second question category, or increased to each proportion of the second question category according to the ratio between multiple proportions of the second question category.

[0111] In a specific example, if the preset proportion corresponding to the first question category level is 80%, the proportion of the first question category corresponding to the first question category level is 90%, the proportion of the second question category corresponding to the other question category levels is 10%, and the preset weighting coefficient is 0.9, then the adjusted proportion of the first question category is 90% × 0.9 = 81%, and the adjusted proportion of the second question category is (90% - 81%) + 10% = 19%.

[0112] The above methods for adjusting the proportions of the first question type and the second question type are merely examples. This embodiment does not limit any other methods for adjusting the proportions of the first question type and the second question type.

[0113] This method avoids having too many target words in the first question category level, allowing for more balanced training of the target user's vocabulary ability corresponding to each target question category level. Furthermore, when the first question category level is relatively difficult, it also avoids having too many difficult questions generated based on the target words corresponding to the first question category level, thereby preventing a decrease in the target user's learning interest.

[0114] Regarding step 102,

[0115] The candidate words are words in the target user's vocabulary list, and the candidate words include at least one of the following: words in a preset vocabulary list (such as CET-4 words, CET-6 words), words collected by the target user, and words selected from the multimedia content browsed by the target user in history.

[0116] In one possible implementation, if the candidate word is selected from the multimedia content browsed in the target user's history, the candidate word can be determined by inputting the multimedia content browsed in the target user's history into a pre-trained language model, and the language model outputs the candidate word.

[0117] Specifically, when inputting multimedia content into a pre-trained language model, if the multimedia content is text information, it can be directly input into the language model; if the multimedia content is audio, the audio can be converted into text information first, and then the text information can be input into the language model; if the multimedia content is an image, text recognition can be performed on the image to obtain the text information contained in the image, and then the text information can be input into the language model; if the multimedia content is video, the text information in the video can be extracted first (including audio converted into text information, text information obtained from image recognition, and text information obtained from subtitle extraction), and then the text information can be input into the language model.

[0118] For any multimedia content, after receiving the multimedia content, the language model can count the frequency of each word in the multimedia content, and then filter out words that appear less frequently in other multimedia content but more frequently in this multimedia content. This avoids filtering out frequently occurring words such as "to," "the," "I," and "you" that have no learning value, and instead filters out nouns, verbs, adjectives, adverbs, etc. that are valuable for the specific theme of the multimedia content.

[0119] In addition, in order to further extract words with learning value from the candidate words, the candidate words can be compared with a preset standard vocabulary, and words that are not in the standard vocabulary can be deleted from the candidate words.

[0120] The target question category level corresponding to the question category ratio is: the question category level in the question category ratio that is not 0. For example, if the question category ratio is: Question Category Level 1: Question Category Level 2: Question Category Level 3: Question Category Level 4: Question Category Level 5 = 5:4:0:0:1, then the target question category level is Question Category Level 1, Question Category Level 2 and Question Category Level 5.

[0121] It is understandable that, since different question levels have different levels of difficulty, users will have a higher correct answer rate for questions generated from words at easier question levels, and a lower correct answer rate for questions generated from words at more difficult question levels. Therefore, in order to accurately predict the correct answer rate of the target user for questions generated from target words according to the question level ratio, the target user's ability coefficient under the question level ratio is used as a score representing the user's vocabulary ability.

[0122] The target historical questions are those of the target question category level from the historical questions that the user has practiced. The historical answer results may include whether each historical question was answered correctly (e.g., correct answer and incorrect answer), the accuracy rate of the historical questions, etc.

[0123] The target user's ability coefficient under the specified question type ratio can be predetermined, or it can be determined at any time before step 102, such as after responding to the word practice instruction. Specifically, it can be calculated based on the historical answer results using a preset first algorithm to obtain the ability coefficient. For example, the target user's ability coefficient under the specified question type ratio can be calculated based on Item Response Theory (IRT); wherein, the IRT model is a statistical mathematical model that can be used to analyze historical answer results to obtain the ability coefficient.

[0124] In one possible implementation, the target user's ability coefficient under the proportion of the question type can also be determined based on the historical answer results of the target historical question and the word difficulty of the candidate words contained in the target historical question.

[0125] It is understandable that a user's ability to answer a more difficult question correctly is more indicative of their vocabulary proficiency than that of an easier question. Therefore, when calculating the ability coefficient, referring to the difficulty of the candidate words can more accurately determine the target user's ability coefficient for the specified question type.

[0126] Specifically, the historical answer results and the word difficulty of the candidate words can be calculated according to a preset second algorithm to obtain the target user's ability coefficient under the question category ratio. For example, the historical answer results and the word difficulty of the candidate words can be input into the IRT model respectively. More specifically, a target matrix can be generated based on the historical answer results and the word difficulty of the candidate words. For example, the first row of the target matrix is ​​the word difficulty of the candidate words, and the second row is whether the target user answered correctly (e.g., 1 for correct answer and 0 for incorrect answer). Then, the IRT model can output the ability coefficient.

[0127] The difficulty of the candidate words can be predetermined, or it can be determined at any time before step 102, such as after responding to the word practice instruction. In one possible implementation, the difficulty of each candidate word can be determined by: determining the difficulty of each candidate word based on the statistical results of answers given by multiple second test users to the questions corresponding to each candidate word.

[0128] Specifically, the answer statistics may include: whether the question corresponding to the candidate word was answered correctly (e.g., correct or incorrect answer), the accuracy rate of the question corresponding to the candidate word, etc. Then, the answer statistics can be calculated using a preset third algorithm to obtain the word difficulty of the candidate word. For example, the answer statistics can be input into an IRT model to obtain the word difficulty of the candidate word.

[0129] Using this method, the difficulty of the candidate words can be accurately estimated based on the actual answers given by the second test user to the questions corresponding to each candidate word.

[0130] It is understandable that users with different educational backgrounds, grades, majors, and professions have different vocabulary abilities and areas of vocabulary exposure. Therefore, different users will have different perceptions of word difficulty. For example, medical words may be very difficult for high school students, while they may be relatively easy for medical students. In order to more accurately estimate the word difficulty of each candidate word for the target user, the second test user can be determined according to the following method:

[0131] In one possible implementation, if the number of first statistical users exceeds a preset number, the plurality of second test users are the first statistical users; if the number of first statistical users does not exceed the preset number, the plurality of second test users are the second statistical users; or they are the first statistical users and the second statistical users, wherein the attribute information of the first statistical users is the same as that of the target user, and the attribute information of the second statistical users is different from that of the target user.

[0132] The attribute information may include at least one of the following: region, occupation, education level, grade, major, and (in the case of being executed by the target application in this embodiment of the disclosure) whether the user is a user within the target application.

[0133] Understandably, since the attribute information of the first statistical user is the same as that of the target user, determining the word difficulty based on the statistical results of the first statistical user's answers more closely reflects the target user's perception of word difficulty. However, if the number of the first statistical users is small, their statistical results are somewhat random and cannot represent the perception of word difficulty among users with the same attribute information as the first statistical user. Therefore, when the number of the first statistical users exceeds the preset number, the word difficulty of the candidate word can be determined based on the statistical results of their answers; when the number of the first statistical users does not exceed the preset number, the word difficulty of the candidate word can be determined based on the statistical results of the answers of the second statistical user (and the first statistical user).

[0134] For example, the target user's attribute information is a medical student, the first statistical user's attribute information is a medical student, the second statistical user's attribute information is a high school student, and the preset number is 5000. If the number of the first statistical users is 4000, the word difficulty of the candidate word is determined based on the statistical results of the high school students' answers (e.g., the word difficulty is 60). If the number of the first statistical users is 5500, the word difficulty of the candidate word is determined based on the statistical results of the first statistical users' answers (e.g., the word difficulty is 50).

[0135] In another example, the difficulty of the candidate words may be determined according to the Global Scale of English (GSE) difficulty, which is based on the statistical results of responses from multiple students (i.e., the second statistical users) in multiple countries, and the first statistical users may be users within the target application.

[0136] Using this method, the difficulty of the candidate words can be accurately estimated based on attribute information, so as to better select target words of appropriate difficulty for the target user and generate target questions.

[0137] In one possible implementation, when determining the target user's correct answer rate for each candidate word based on the target user's ability coefficient under the question category ratio, the correct answer rate for each candidate word can be determined based on the target user's ability coefficient under the question category ratio, the word difficulty of each candidate word, and a preset target parameter; wherein, the target parameter characterizes the degree of influence of the historical browsing status of each candidate word on the correct answer rate, and the target parameter is determined based on the statistical results of multiple first test users' answers to browsed words and the statistical results of their answers to unbrowsed words.

[0138] For example, the correct answer rate of the candidate words can be calculated using the following formula:

[0139] P = σ(θ - b + δ)

[0140] Wherein, P represents the correct answer rate of the candidate word, θ represents the ability coefficient of the target user under the proportion of the question type, b represents the word difficulty of the candidate word, and δ represents the target parameter.

[0141] It is understandable that the target user's vocabulary ability and the difficulty of the candidate words will affect the target user's ability to answer questions corresponding to the candidate words. In addition, the candidate words are usually words that the user has browsed (such as those they have learned or those they have viewed from multimedia content). If the target user has browsed the candidate word, the user will have a higher chance of answering questions corresponding to the candidate word correctly. Therefore, introducing the target parameter into the calculation can make the correct answer rate of the candidate words more accurate.

[0142] Specifically, when determining the target parameter, statistical results of responses from multiple first test users to multiple first words (i.e., words already viewed by the first test users) and statistical results of responses from multiple first test users to multiple second words (i.e., words not viewed by the first test users) can be collected, and the target parameter can be calculated using a fourth preset algorithm; or, statistical results of responses from a first group of users among the multiple first test users to the first word (i.e., words already viewed by the first group of users) and statistical results of responses from a second group of users (excluding the first group of users) to the first word (i.e., words not viewed by the second group of users) can be collected, and the target parameter can be calculated using a fourth preset algorithm.

[0143] This method combines multiple factors, such as the user's vocabulary ability, the difficulty of the candidate words themselves, and the degree to which the user is more likely to answer correctly with words they have browsed, to comprehensively determine the correct answer rate of the candidate words, thereby improving the accuracy of the calculated correct answer rate.

[0144] Regarding step 103,

[0145] The number of words to be practiced carried in the word practice instruction can be set by the target user, sent by another user or server, or be a preset default value. In a specific application scenario, when the target user creates a word practice plan, they can select the number of words to practice each day, such as 30. Then, after triggering the "Start Practice" button, a word practice instruction carrying the specified number of words to be practiced is generated.

[0146] Specifically, step 103 can be divided into the following steps A1 to A2:

[0147] A1. Based on the proportion of the question types and the number of word exercises, determine the number of word exercises for each target question type level in the target words.

[0148] For example, if the question category ratio is: Question Category Level 1: Question Category Level 2: Question Category Level 3 = 1:2:3, and the number of word exercises is 30, then the number of word exercises corresponding to Question Category Level 1 is 5, the number of word exercises corresponding to Question Category Level 2 is 10, and the number of word exercises corresponding to Question Category Level 3 is 15.

[0149] A2. For any candidate words corresponding to a mastery level, select the candidate words whose correct answer rate is closest to the target number of correct answer rates, and use them as the target words corresponding to that mastery level; wherein, the target number is the number of words practiced at the same target level as that mastery level.

[0150] For example, if the preset correct answer rate is 75%, the candidate words corresponding to question category level 1 and the correct answer rates corresponding to the candidate words are as follows: word 1 (90%), word 2 (85%), word 3 (80%), word 4 (76%), word 5 (73%), word 6 (60%), and word 7 (50%). The number of practice words corresponding to question category level 1 is 3. Then the target words corresponding to question category level 1 are word 4, word 5, and word 3, whose correct answer rates are closest to the preset correct answer rate.

[0151] In one possible implementation, when the candidate words are selected from the multimedia content browsed in the target user's history (and actively collected by the target user), when selecting target words that meet the question category ratio from multiple candidate words based on the number of word practice sessions carried by the word practice instruction, the correct answer rate, and the target user's mastery level of the candidate words, the selection can be based on the number of word practice sessions carried by the word practice instruction, the correct answer rate, the target user's mastery level of the candidate words, and the target time corresponding to each candidate word; wherein, the target time corresponding to each candidate word includes the selection time corresponding to each candidate word and / or the historical practice time corresponding to each candidate word.

[0152] The selection time corresponding to each candidate word can be the time when the candidate word was selected from the multimedia content browsed in the target user's history (in the case that the candidate word is a word actively collected by the target user, the selection time is the time when the word was collected), and the historical practice time corresponding to each candidate word can be the last practice time of the candidate word that has been practiced.

[0153] Specifically, the target word can be determined by following the steps B1 to B4:

[0154] B1. Based on the proportion of the question types and the number of word exercises, determine the number of word exercises for each target question type level in the target word.

[0155] The specific process here is the same as step A1 above, and will not be repeated here.

[0156] B2. For any candidate word corresponding to a mastery level, determine the preset correct answer rate range to which the candidate word corresponding to that mastery level belongs.

[0157] For example, the correct answer rate ranges are 10% to 20%, 20% to 30%, 30% to 40%, etc. If the correct answer rate for word 1 is 25%, then word 1 belongs to the 20% to 30% correct answer rate range.

[0158] B3. Sort each candidate word according to the correct answer rate interval to which it belongs, and sort the candidate words in each correct answer rate interval according to the corresponding target time.

[0159] Specifically, the correct answer rate intervals can be sorted first according to a preset correct answer rate, and then the candidate words can be sorted according to the sorting of the correct answer rate intervals, such that the correct answer rate intervals closer to the preset correct answer rate are ranked higher. For example, if the preset correct answer rate is 75%, then the correct answer rate intervals of 70% to 80% are ranked before those of 50% to 60%.

[0160] Then, the candidate words within each correct answer rate interval are sorted according to the target time. For example, if the target time is a selected time, the time interval between the selected time and the current time is sorted from smallest to largest. For example, candidate words with a selected time of January 5th are ranked before candidate words with a selected time of January 4th. This allows the target user to prioritize practicing the words that were added as candidate words first.

[0161] When the target time is a historical practice time, the time interval between the historical practice time and the current time is sorted from largest to smallest. For example, candidate words with a historical practice time of May 20th are ranked ahead of candidate words with a historical practice time of August 5th. This allows the target user to prioritize practicing words with a longer historical practice time from the current time.

[0162] In one possible implementation, for any candidate word corresponding to a mastery level, if the candidate word is a word that the target user has already practiced, it can be sorted according to the historical practice time; if the candidate word is a word that the target user has not practiced, it can be sorted according to the selected time; if the candidate word includes both unpracticed and practiced words, the candidate word is sorted according to a preset sorting method between unpracticed and practiced words, and the unpracticed words are sorted according to the selected time, and the practiced words are sorted according to the historical practice time. For example, the unpracticed words are first placed before the practiced words, then the unpracticed words are sorted according to the selected time, and the practiced words are sorted according to the historical practice time.

[0163] B4. Select the target number of candidate words from the sorted candidate words and use them as the target words corresponding to the mastery level; wherein, the target number is the number of word practice sessions corresponding to the same mastery level.

[0164] Regarding step 104,

[0165] Specifically, when generating the target question corresponding to the target word, the target question can be generated based on the target word and a preset question template, or the target question corresponding to the target word can be searched from a preset question bank.

[0166] For example, if the preset question template is "What is the Chinese definition of ____", and the target word is "apple", then the generated target question will be "What is the Chinese definition of apple".

[0167] After generating the target question, the target question can be displayed to the target user for the target user to answer, and the answer result of the target user can be collected after the target user answers.

[0168] In one possible implementation, when generating the target question corresponding to the target word, the following steps C1 to C3 may also be performed:

[0169] C1. Determine at least one candidate question type corresponding to the mastery level of the target word.

[0170] Specifically, at least one candidate question type can be pre-set for each question category level, and then at least one candidate question type can be queried according to the mastery level of the question category. The question category level and the corresponding candidate question type can be exemplified as shown in Table 2 below.

[0171]

[0172] Table 2

[0173] C2. Based on the priority of the at least one candidate question type, determine the target question type corresponding to the target word from the at least one candidate question type.

[0174] The priority of each candidate question type is exemplified in Table 2, such as the priority of selecting the meaning of a word from a video being 1, and the priority of selecting the meaning of a word from a video being 2.

[0175] Specifically, the priorities of at least one candidate question type can be compared, and the candidate question type with the highest priority can be selected as the target question type.

[0176] Here, if the candidate question type with the highest priority cannot generate the target question, the candidate question type with the next lower priority adjacent to the current priority can be used as the target question type. For example, if the mastery level of the target word is level 3, and the candidate question type with the highest priority is "listen and choose the meaning", but there is no audio corresponding to the target word in the target application, only the audio video corresponding to the target word, then listening and choosing the meaning from the audio or video can be used as the target question type.

[0177] C3. Generate the target question based on the target word and the target question type.

[0178] Specifically, the target question can be generated based on the target word and the preset question template corresponding to the target question type.

[0179] Using this method, question types with better practice effects can be selected based on priority, so that the generated target questions can better train the user's vocabulary ability.

[0180] The target questions can typically be multiple choice questions, fill-in-the-blank questions, reading aloud questions, essay questions, etc. When generating multiple choice questions, in addition to determining the stem of the target question, it is also necessary to determine the options of the target question.

[0181] Therefore, in one possible implementation, when generating the target question based on the target word and the target question type, if the target question type is a multiple-choice question, the target distractor word corresponding to the target word can be determined; then, based on the target distractor word, the target question type, and the target word, the target question corresponding to the target word can be generated.

[0182] Specifically, after determining the target question type corresponding to the target word, it can be determined whether the target question type is a multiple-choice question. If so, the question stem of the target question can be generated based on the question template corresponding to the multiple-choice question (or the target question type) and the target word. Additionally, the target distractor words corresponding to the target word can be determined, and the options in the target question can be generated. Then, the target question corresponding to the target word can be generated based on the target word, the question stem, and the options. The target distractor words corresponding to the target word can be the target distractor words corresponding to the target word under the target question type.

[0183] For example, if the target word is "apple" and the question template is "What is the Chinese meaning of ___?", then the question stem corresponding to the target word is "What is the Chinese meaning of apple?". If the target distractor words corresponding to "apple" are "orange" and "banana", then the options "A. orange", "B. banana", and "C. apple" can be generated. Then, the target question generated based on the question stem and the options is "What is the Chinese meaning of apple? A. orange; B. banana; C. apple".

[0184] Using this method, multiple-choice questions corresponding to the target question can be generated automatically, without the need for manual pre-setting of multiple-choice questions for each word, saving human resources and improving the efficiency of question generation.

[0185] In one possible implementation, when determining the target interference word corresponding to the target word, the preset candidate interference words corresponding to the target word and the historical words contained in the target user's historical questions can be determined first; wherein, the historical words include historical practice words and / or historical interference words; and then the target interference words other than the historical words are filtered out from the candidate interference words.

[0186] Specifically, the historical questions are those the target user has already practiced, the historical practice words are words the user has already practiced used to generate the historical questions, and the historical distractor words are options from the historical multiple-choice questions. Each target word has multiple pre-set candidate distractor words. The target word can be directly queried for its corresponding candidate distractor words. When determining historical words, the target user's historical questions can be obtained first, and then the historical words contained within those questions can be determined. Alternatively, the historical words contained in each question can be determined after the target user completes it, meaning the pre-determined historical words can be directly obtained when determining the historical words. Then, the historical words from the candidate distractor words are excluded, and a preset number of distractor words are randomly selected as the target distractor words.

[0187] By using this method, since the target user has already mastered the words that have been practiced or have appeared as options, these words can be excluded, and words that are less familiar to the target user can be selected as options, so that the target user can learn more new words and improve learning efficiency.

[0188] Understandably, for any candidate word, in order to enable users to master the various vocabulary skills related to that candidate word, an assessment can be conducted using questions at each question category level corresponding to that candidate word. Therefore, before the candidate word is assessed, its question category level can be set to an initial level (such as question category level 1 in Table 1). Then, after the target user completes the target questions corresponding to the candidate word (which is now the target word), the mastery level of the target word can be updated. The following is a specific update method:

[0189] In one possible implementation, after displaying the target question to the target user, the target user's target answer to the target question can be obtained; then, based on the target answer and a preset question category level update rule, the mastery question category level of the target word is updated; wherein, the question category level update rule is used to indicate the method for changing the mastery question category level of each target word in the case of correct or incorrect answers to the corresponding target question.

[0190] Specifically, the question category level update rule is as follows: if the target question is answered correctly, the mastery level of the target word is updated to a higher level of difficulty, such as upgrading question category level 2 to question category level 3. If the target question is answered incorrectly, the mastery level of the target word is updated to a lower level of difficulty, such as downgrading question category level 5 to question category level 4. Here, since questions at question category level 2 are relatively simple and easy to master, if the mastery level of the target word is at the first mastery level, the mastery level of the target word may not be updated when the target question is answered incorrectly, allowing users to practice more questions at the first mastery level.

[0191] Using this method, the target user's mastery level of the target word can be updated based on their target answer results. This allows for adjustments to the target question types generated for the next test of the target word, as well as the proportion of target questions generated for each question type, based on the target user's mastery of the target word and their ability to answer questions at each question type level. In this way, the questions for the next practice session can be adjusted in a personalized manner based on the user's answer results each time.

[0192] In one possible implementation, when the mastery level of the target word is the second mastery level, and the target answer is correct, the target word can be removed from the candidate words. The second mastery level can be the highest among the preset multiple question category levels.

[0193] It is understandable that if the target user's mastery level of the target word reaches the second mastery level and the target user answers the target question correctly, it indicates that the target user has mastered the highest difficulty level of the target word. At this point, it can be considered that the target user has mastered all the vocabulary skills related to the target word, and further training on the target word is no longer necessary.

[0194] Finally, this disclosure also provides an overall flow of a question generation method, such as... Figure 2 As shown, see steps 201 to 207 below:

[0195] Step 201: In response to the word practice instruction, based on the mastery level of each candidate word, determine the proportion of question categories corresponding to multiple preset question categories.

[0196] Step 202: Based on the statistical results of the answers given by multiple second test users to the questions corresponding to each candidate word, determine the word difficulty of each candidate word;

[0197] Step 203: Based on the historical answers to the target historical questions and the word difficulty of the candidate words contained in the target historical questions, determine the ability coefficient of the target user under the proportion of the question type;

[0198] Step 204: Based on the target user's ability coefficient under the question category ratio, the word difficulty of each candidate word, and the preset target parameters, determine the target user's correct answer rate for each candidate word; wherein, the target parameters characterize the degree of influence of the historical browsing status of each candidate word on the correct answer rate, and the target parameters are determined based on the statistical results of multiple first test users' answers to browsed words and the statistical results of their answers to unbrowsed words;

[0199] Step 205: Based on the number of words practiced in the word practice instruction, the correct answer rate, and the target user's mastery level of the candidate words, select target words that meet the question category ratio from multiple candidate words;

[0200] Step 206: Determine at least one candidate question type corresponding to the mastery level of the target word; based on the priority of the at least one candidate question type, determine the target question type corresponding to the target word from the at least one candidate question type;

[0201] Step 207: If the target question type is a multiple-choice question, determine the target distractor word corresponding to the target word; based on the target distractor word, the target question type, and the target word, generate the target question corresponding to the target word.

[0202] The question generation method provided in this embodiment can first determine the question category ratios corresponding to multiple preset question category levels after responding to a word practice instruction. Since different question category levels are used to test different word abilities, the target user's correct answer rate for each candidate word can be accurately determined based on the target user's ability coefficient under the question category ratios. Then, based on the number of words practiced in the word practice instruction, the correct answer rate, and the target user's mastery level of the candidate words, target words that meet the question category ratios are selected from the multiple candidate words. In this way, since the difficulty of the target words is suitable for the target user's current level, the target questions generated based on the target words are also of moderate difficulty for the target user, thereby improving the user's learning efficiency and learning experience.

[0203] Those skilled in the art will understand that, in the above-described method of the specific implementation, the order in which each step is written does not imply a strict execution order and does not constitute any limitation on the implementation process. The specific execution order of each step should be determined by its function and possible internal logic.

[0204] Based on the same inventive concept, this disclosure also provides a question generation device corresponding to the question generation method. Since the principle of the device in this disclosure for solving the problem is similar to the question generation method described above, the implementation of the device can refer to the implementation of the method, and the repeated parts will not be described again.

[0205] Reference Figure 3 The diagram shown is an architectural schematic of a question generation device provided in an embodiment of this disclosure. The device includes: a first determining module 301, a second determining module 302, a filtering module 303, and a generation module 304; wherein,

[0206] The first determining module 301 is used to respond to the word practice instruction and determine the proportion of question types corresponding to multiple preset question types and levels; wherein, different question types and levels are used to test different word abilities;

[0207] The second determining module 302 is used to determine the correct answer rate of the target user for each candidate word based on the target user's ability coefficient under the question category ratio; wherein, the target user's ability coefficient under the question category ratio is determined based on the target user's historical answer results for target historical questions of the target question category level corresponding to the question category ratio;

[0208] The filtering module 303 is used to filter target words that meet the question category ratio from multiple candidate words based on the number of words practiced in the word practice instruction, the correct answer rate, and the target user's mastery level of the candidate words.

[0209] The generation module 304 is used to generate the target question corresponding to the target word and display the target question through the target user terminal.

[0210] In one possible implementation, the first determining module 301, when determining the proportion of question types corresponding to multiple preset question type levels, is used to:

[0211] Based on the mastery level of each candidate word, the proportion of each question type corresponding to a number of preset question types is determined.

[0212] In one possible implementation, after determining the proportion of question types corresponding to multiple preset question type levels, the first determining module 301 is further configured to:

[0213] If the question category ratio includes a first question category level and other question category levels besides the first question category level, and the first question category ratio corresponding to the first question category level exceeds a preset ratio, then the first question category ratio corresponding to the first question category level and the second question category ratio corresponding to the other question category levels are adjusted.

[0214] In one possible implementation, the device is also used to determine the target user's ability coefficient under the proportion of the question type according to the following method:

[0215] Based on the historical answers to the target historical questions and the word difficulty of the candidate words contained in the target historical questions, the ability coefficient of the target user under the proportion of the question type is determined.

[0216] In one possible implementation, the second determining module 302, when determining the target user's correct answer rate for each candidate word based on the target user's ability coefficient under the question category ratio, is used to:

[0217] Based on the target user's ability coefficient under the question category ratio, the word difficulty of each candidate word, and the preset target parameters, the target user's correct answer rate for each candidate word is determined;

[0218] The target parameter represents the degree of influence of the historical browsing status of each candidate word on the correct answer rate. The target parameter is determined based on the statistical results of the answers of multiple first test users to browsed words and the statistical results of the answers to unbrowsed words.

[0219] In one possible implementation, the device is also used to determine the word difficulty of each of the candidate words according to the following method:

[0220] Based on the statistical results of multiple second test users' answers to the questions corresponding to each candidate word, the word difficulty of each candidate word is determined.

[0221] In one possible implementation, if the number of first statistical users exceeds a preset number, the plurality of second test users are the first statistical users; if the number of first statistical users does not exceed the preset number, the plurality of second test users are the second statistical users; or they are the first statistical users and the second statistical users, wherein the attribute information of the first statistical users is the same as that of the target user, and the attribute information of the second statistical users is different from that of the target user.

[0222] In one possible implementation, the candidate words are selected from the multimedia content that the target user has browsed in the past;

[0223] The filtering module 303, when filtering target words that meet the question category ratio from multiple candidate words based on the number of words practiced in the word practice instruction, the correct answer rate, and the target user's mastery level of the candidate words, is used to:

[0224] Based on the number of words practiced in the word practice instruction, the correct answer rate, the target user's mastery level of the candidate words, and the target time corresponding to each candidate word, target words that meet the question type ratio are selected from multiple candidate words; wherein, the target time corresponding to each candidate word includes the selected time corresponding to each candidate word and / or the historical practice time corresponding to each candidate word.

[0225] In one possible implementation, the generation module 304, when generating the target question corresponding to the target word, is used to:

[0226] Identify at least one candidate question type corresponding to the mastery level of the target word;

[0227] Based on the priority of the at least one candidate question type, the target question type corresponding to the target word is determined from the at least one candidate question type;

[0228] Based on the target words and the target question type, the target question is generated.

[0229] In one possible implementation, the generation module 304, when generating the target question based on the target word and the target question type, is used to:

[0230] When the target question type is multiple choice, determine the target distractor word corresponding to the target word;

[0231] Based on the target distractor word, the target question type, and the target word, generate the target question corresponding to the target word.

[0232] In one possible implementation, the generation module 304, when determining the target interference word corresponding to the target word, is used to:

[0233] Determine the preset candidate distractor words corresponding to the target word, as well as the historical words contained in the target user's historical questions; wherein, the historical words include historical practice words and / or historical distractor words;

[0234] From the candidate interference words, select the target interference words other than the historical words.

[0235] In one possible implementation, after displaying the target question to the target user, the device is further configured to:

[0236] Obtain the target user's target answer to the target question;

[0237] Based on the target answer results and the preset question category level update rules, the mastery question category level of the target words is updated; wherein, the question category level update rules are used to indicate the method for changing the mastery question category level of each target word when the corresponding target question is answered correctly or incorrectly.

[0238] The processing flow of each module in the device and the interaction flow between each module can be referred to the relevant descriptions in the above method embodiments, and will not be detailed here.

[0239] Based on the same technical concept, this disclosure also provides a computer device. (See also...) Figure 4 The diagram shows the structure of a computer device 400 provided in this embodiment, including a processor 401, a memory 402, and a bus 403. The memory 402 stores execution instructions and includes main memory 4021 and external memory 4022. The main memory 4021, also called internal memory, is used to temporarily store computational data in the processor 401 and data exchanged with external memory 4022 such as a hard disk. The processor 401 exchanges data with the external memory 4022 through the main memory 4021. When the computer device 400 is running, the processor 401 and the memory 402 communicate through the bus 403, causing the processor 401 to execute the following instructions:

[0240] In response to the vocabulary practice instruction, the system determines the proportion of each of the multiple preset question levels; different question levels are used to assess different vocabulary abilities.

[0241] Based on the target user's ability coefficient under the specified question category ratio, the target user's correct answer rate for each candidate word is determined; wherein, the target user's ability coefficient under the specified question category ratio is determined based on the target user's historical answer results for target question categories at the target question category level corresponding to the specified question category ratio;

[0242] Based on the number of words practiced in the word practice instruction, the correct answer rate, and the target user's mastery level of the candidate words by question type, target words that meet the question type ratio are selected from multiple candidate words;

[0243] Generate the target question corresponding to the target word and display the target question through the target user terminal.

[0244] In one possible implementation, the instruction executed by processor 401, which determines the proportion of question types corresponding to multiple preset question type levels, includes:

[0245] Based on the mastery level of each candidate word, the proportion of each question type corresponding to a number of preset question types is determined.

[0246] In one possible implementation, after determining the proportion of question types corresponding to multiple preset question type levels in the instructions executed by processor 401, the method further includes:

[0247] If the question category ratio includes a first question category level and other question category levels besides the first question category level, and the first question category ratio corresponding to the first question category level exceeds a preset ratio, then the first question category ratio corresponding to the first question category level and the second question category ratio corresponding to the other question category levels are adjusted.

[0248] In one possible implementation, the instructions executed by processor 401 further include determining the target user's ability coefficient under the question category ratio according to the following method:

[0249] Based on the historical answers to the target historical questions and the word difficulty of the candidate words contained in the target historical questions, the ability coefficient of the target user under the proportion of the question type is determined.

[0250] In one possible implementation, the instruction executed by processor 401, which determines the target user's correct answer rate for each candidate word based on the target user's ability coefficient under the question category ratio, includes:

[0251] Based on the target user's ability coefficient under the question category ratio, the word difficulty of each candidate word, and the preset target parameters, the target user's correct answer rate for each candidate word is determined;

[0252] The target parameter represents the degree of influence of the historical browsing status of each candidate word on the correct answer rate. The target parameter is determined based on the statistical results of the answers of multiple first test users to browsed words and the statistical results of the answers to unbrowsed words.

[0253] In one possible implementation, the instructions executed by processor 401 further include determining the word difficulty of each of the candidate words according to the following method:

[0254] Based on the statistical results of multiple second test users' answers to the questions corresponding to each candidate word, the word difficulty of each candidate word is determined.

[0255] In one possible implementation, in the instructions executed by the processor 401, if the number of first statistical users exceeds a preset number, the plurality of second test users are the first statistical users; if the number of first statistical users does not exceed the preset number, the plurality of second test users are second statistical users; or they are the first statistical user and the second statistical user, wherein the attribute information of the first statistical user is the same as that of the target user, and the attribute information of the second statistical user is different from that of the target user.

[0256] In one possible implementation, the candidate words in the instructions executed by the processor 401 are selected from the multimedia content that the target user has browsed in the past.

[0257] The step of selecting target words that meet the question category ratio from multiple candidate words based on the number of words practiced in the word practice instruction, the correct answer rate, and the target user's mastery level of the candidate words includes:

[0258] Based on the number of words practiced in the word practice instruction, the correct answer rate, the target user's mastery level of the candidate words, and the target time corresponding to each candidate word, target words that meet the question type ratio are selected from multiple candidate words; wherein, the target time corresponding to each candidate word includes the selected time corresponding to each candidate word and / or the historical practice time corresponding to each candidate word.

[0259] In one possible implementation, the instructions executed by processor 401, including generating the target question corresponding to the target word, include:

[0260] Identify at least one candidate question type corresponding to the mastery level of the target word;

[0261] Based on the priority of the at least one candidate question type, the target question type corresponding to the target word is determined from the at least one candidate question type;

[0262] Based on the target words and the target question type, the target question is generated.

[0263] In one possible implementation, the instructions executed by processor 401, including generating the target question based on the target word and the target question type, include:

[0264] When the target question type is multiple choice, determine the target distractor word corresponding to the target word;

[0265] Based on the target distractor word, the target question type, and the target word, generate the target question corresponding to the target word.

[0266] In one possible implementation, the instructions executed by processor 401, including determining the target interference word corresponding to the target word, include:

[0267] Determine the preset candidate distractor words corresponding to the target word, as well as the historical words contained in the target user's historical questions; wherein, the historical words include historical practice words and / or historical distractor words;

[0268] From the candidate interference words, select the target interference words other than the historical words.

[0269] In one possible implementation, after displaying the target question through the target user terminal, the method further includes the following instructions executed by processor 401:

[0270] Obtain the target user's target answer to the target question;

[0271] Based on the target answer results and the preset question category level update rules, the mastery question category level of the target words is updated; wherein, the question category level update rules are used to indicate the method for changing the mastery question category level of each target word when the corresponding target question is answered correctly or incorrectly.

[0272] This disclosure also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, performs the steps of the question generation method described in the above method embodiments. The storage medium can be a volatile or non-volatile computer-readable storage medium.

[0273] This disclosure also provides a computer program product carrying program code. The program code includes instructions that can be used to execute the steps of the question generation method described in the above method embodiments. For details, please refer to the above method embodiments, which will not be repeated here.

[0274] The aforementioned computer program product can be implemented through hardware, software, or a combination thereof. In one optional embodiment, the computer program product is specifically embodied in a computer storage medium; in another optional embodiment, the computer program product is specifically embodied in a software product, such as a software development kit (SDK), etc.

[0275] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems and devices described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here. In the several embodiments provided in this disclosure, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division; in actual implementation, there may be other division methods. Furthermore, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Another point is that the displayed or discussed mutual coupling or direct coupling or communication connection may be through some communication interfaces; the indirect coupling or communication connection of devices or units may be electrical, mechanical, or other forms.

[0276] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0277] In addition, the functional units in the various embodiments of this disclosure can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0278] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a processor-executable, non-volatile, computer-readable storage medium. Based on this understanding, the technical solution of this disclosure, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this disclosure. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0279] Finally, it should be noted that the above-described embodiments are merely specific implementations of this disclosure, used to illustrate the technical solutions of this disclosure, and not to limit it. The protection scope of this disclosure is not limited thereto. Although this disclosure has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments, or make equivalent substitutions for some of the technical features, within the scope of the technology disclosed in this disclosure. Such modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this disclosure, and should all be covered within the protection scope of this disclosure. Therefore, the protection scope of this disclosure should be determined by the protection scope of the claims.

Claims

1. A method for generating questions, characterized in that, include: In response to a word practice instruction, the proportion of question types corresponding to multiple preset question levels is determined; wherein, different question levels are used to test different word abilities, and the proportion of question types represents the ratio of the number of target words to be practiced corresponding to the multiple preset question levels; Based on the target user's ability coefficient under the specified question category ratio, the target user's correct answer rate for each candidate word is determined; wherein, the target user's ability coefficient under the specified question category ratio is determined based on the target user's historical answer results for target question category level corresponding to the specified question category ratio; Based on the number of words practiced in the word practice instruction, the correct answer rate, and the target user's mastery level of the candidate words by question type, target words that meet the question type ratio are selected from multiple candidate words; Generate target questions corresponding to the target words and display the target questions to the target users. The step of selecting target words that meet the question category ratio from multiple candidate words based on the number of words practiced in the word practice instruction, the correct answer rate, and the target user's mastery level of the candidate words includes: Based on the proportion of question types and the number of word exercises, determine the number of word exercises for each target question type level in the target word; For any candidate word corresponding to a mastery level, select the candidate word whose correct answer rate is closest to the target number of the preset correct answer rate, and use it as the target word corresponding to the mastery level; wherein, the target number is the number of word practice sessions corresponding to the same mastery level as the mastery level.

2. The method according to claim 1, characterized in that, The step of determining the proportion of question types corresponding to multiple preset question type levels includes: Based on the mastery level of each candidate word, the proportion of each question type corresponding to a number of preset question types is determined.

3. The method according to claim 1 or 2, characterized in that, After determining the proportion of question types corresponding to multiple preset question level levels, the method further includes: If the question category ratio includes a first question category level and other question category levels besides the first question category level, and the first question category ratio corresponding to the first question category level exceeds a preset ratio, then the first question category ratio corresponding to the first question category level and the second question category ratio corresponding to the other question category levels are adjusted.

4. The method according to claim 1, characterized in that, The method further includes determining the target user's ability coefficient for the proportion of the question type according to the following method: Based on the historical answers to the target historical questions and the word difficulty of the candidate words contained in the target historical questions, the ability coefficient of the target user under the proportion of the question type is determined.

5. The method according to claim 1, characterized in that, The determination of the target user's correct answer rate for each candidate word based on the target user's ability coefficient under the proportion of the question type includes: Based on the target user's ability coefficient under the question category ratio, the word difficulty of each candidate word, and the preset target parameters, the target user's correct answer rate for each candidate word is determined; The target parameter represents the degree of influence of the historical browsing status of each candidate word on the correct answer rate. The target parameter is determined based on the statistical results of the answers of multiple first test users to browsed words and the statistical results of the answers to unbrowsed words.

6. The method according to claim 5, characterized in that, The method further includes determining the word difficulty of each candidate word according to the following method: Based on the statistical results of multiple second test users' answers to the questions corresponding to each candidate word, the word difficulty of each candidate word is determined.

7. The method according to claim 6, characterized in that, If the number of first statistical users exceeds a preset number, the plurality of second test users are the first statistical users. If the number of first statistical users does not exceed the preset number, the plurality of second test users are the second statistical users, or the first statistical user and the second statistical user. The attribute information of the first statistical user is the same as that of the target user, and the attribute information of the second statistical user is different from that of the target user.

8. The method according to claim 1, characterized in that, The candidate words are selected from the multimedia content that the target user has browsed in the past; The step of selecting candidate words for any mastery level, and choosing the candidate words whose correct answer rate is closest to the preset target number of correct answer rates, as the target words for the mastery level, includes: Based on the number of words practiced in the word practice instruction, the correct answer rate, the target user's mastery level of the candidate words, and the target time corresponding to each candidate word, target words that meet the question type ratio are selected from multiple candidate words; wherein, the target time corresponding to each candidate word includes the selected time corresponding to each candidate word and / or the historical practice time corresponding to each candidate word.

9. The method according to claim 1, characterized in that, The generation of the target question corresponding to the target word includes: Identify at least one candidate question type corresponding to the mastery level of the target word; Based on the priority of the at least one candidate question type, the target question type corresponding to the target word is determined from the at least one candidate question type; Based on the target words and the target question type, the target question is generated.

10. The method according to claim 9, characterized in that, The process of generating the target question based on the target word and the target question type includes: When the target question type is multiple choice, determine the target distractor word corresponding to the target word; Based on the target distractor word, the target question type, and the target word, generate the target question corresponding to the target word.

11. The method according to claim 10, characterized in that, Determining the target interference word corresponding to the target word includes: Determine the preset candidate distractor words corresponding to the target word, as well as the historical words contained in the target user's historical questions; wherein, the historical words include historical practice words and / or historical distractor words; From the candidate interference words, select the target interference words other than the historical words.

12. The method according to claim 1, characterized in that, After displaying the target question to the target user, the method further includes: Obtain the target user's target answer to the target question; Based on the target answer results and the preset question category level update rules, the mastery question category level of the target words is updated; wherein, the question category level update rules are used to indicate the method for changing the mastery question category level of each target word when the corresponding target question is answered correctly or incorrectly.

13. A question generation device, characterized in that, include: The first determining module is used to respond to a word practice instruction and determine the proportion of question types corresponding to multiple preset question type levels; wherein, different question type levels are used to test different word abilities, and the question type proportion represents the ratio of the number of target words to be practiced corresponding to the multiple preset question type levels; The second determining module is used to determine the target user's correct answer rate for each candidate word based on the target user's ability coefficient under the question category ratio; wherein, the target user's ability coefficient under the question category ratio is determined based on the target user's historical answer results for target historical questions of the target question category level corresponding to the question category ratio; The filtering module is used to filter target words that meet the question category ratio from multiple candidate words based on the number of words practiced in the word practice instruction, the correct answer rate, and the target user's mastery level of the candidate words. The generation module is used to generate target questions corresponding to the target words and display the target questions through the target user's terminal. The step of selecting target words that meet the question category ratio from multiple candidate words based on the number of words practiced in the word practice instruction, the correct answer rate, and the target user's mastery level of the candidate words includes: Based on the proportion of question types and the number of word exercises, determine the number of word exercises for each target question type level in the target word; For any candidate word corresponding to a mastery level, select the candidate word whose correct answer rate is closest to the target number of the preset correct answer rate, and use it as the target word corresponding to the mastery level; wherein, the target number is the number of word practice sessions corresponding to the same mastery level as the mastery level.

14. A computer device, characterized in that, include: The computer device includes a processor, a memory, and a bus. The memory stores machine-readable instructions executable by the processor. When the computer device is running, the processor communicates with the memory via the bus. When the machine-readable instructions are executed by the processor, they perform the steps of the question generation method as described in any one of claims 1 to 12.

15. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, performs the steps of the question generation method as described in any one of claims 1 to 12.

Images