Information recommendation method and device, system, electronic equipment and readable storage medium

Abstract

The present disclosure provides an information recommendation method and device, relates to the technical field of information recommendation, and particularly relates to the technical fields of natural language processing, statistics, machine learning and large language models. The specific implementation scheme is as follows: based on a data set when a learning object completes an initial course, a set of association rules is determined and sent; the interest degree of a user for the set of association rules is obtained; the set of association rules is matched with rules in a knowledge base to determine a matching result of the set of association rules; based on the interest degree, the matching result and the set of association rules, information tuples in the knowledge base are updated; based on the information tuples in the knowledge base, a recommended course is determined, and the recommended course is sent to the learning object.

Description

TECHNICAL FIELD

[0001] The present disclosure belongs to the technical field of information recommendation, and particularly relates to the technical fields of natural language processing, machine learning, and large language models. In particular, an information recommendation method and device, an information recommendation system, an electronic device, and a computer readable storage medium. BACKGROUND

[0002] Data mining techniques are increasingly applied in online learning and web-based adaptive education systems. In online learning, data mining is applied to discover useful information that can be directly used by course takers or authors.

[0003] Applying data mining in online learning, particularly an author-centered approach, aims to improve courses, involving a series of obstacles that need to be overcome. Data mining tools usually focus more on power and flexibility than on simplicity and ease of use. Most current data mining tools are too complex for educators, and their features do not cover the range that educators might need. On the one hand, there are many e-learning and web-based adaptive courses that can apply data mining, which are influenced by three key aspects: first, the knowledge domain covered by the course; second, the course level (university, middle school or primary school level, special education or other types of courses); and finally, the difficulty level of the course, i.e., whether it is a basic or beginner, intermediate, advanced or expert course. On the other hand, the range of results that can be obtained according to these factors is very broad, which means that it is quite difficult to find a generalizable and repeatable pattern that can be applied to any type of course. In addition, education datasets are usually small if compared with databases used in other data mining fields, such as e-commerce applications, which involve thousands of customers. Therefore, applying specific filtering parameters to data mining can cause problems in finding association rules in small databases, in which the initial information is not enough to build a model to infer future recommended information. SUMMARY

[0004] The present disclosure provides an information recommendation method and device, an information recommendation system, an electronic device, and a computer readable storage medium.

[0005] According to a first aspect, an information recommendation method is provided, which comprises: determining and sending a set of association rules based on a dataset of a learning object completing an initial course; obtaining an interest degree of a user on the set of association rules; matching the set of association rules with rules in a knowledge base to determine a matching result of the set of association rules; updating information tuples in the knowledge base based on the interest degree, the matching result and the set of association rules; and determining a recommended course based on the information tuples in the knowledge base and sending the recommended course to the learning object.

[0006] According to a second aspect, an information recommendation apparatus is provided, comprising: a determination unit configured to determine and send a set of association rules based on a data set when a learning object finishes an initial course; an acquisition unit configured to acquire an interest degree of a user to the set of association rules; a matching unit configured to match the set of association rules with rules in a knowledge base, and determine a matching result of the set of association rules; an update unit configured to update information tuples in the knowledge base based on the interest degree, the matching result and the set of association rules; and a recommendation unit configured to determine a recommended course based on the information tuples in the knowledge base, and send the recommended course to the learning object.

[0007] According to a third aspect, an information recommendation system is provided, comprising: a client and a server, the client is configured to determine and send a set of association rules based on a data set when a learning object finishes an initial course; acquire an interest degree of a user to the set of association rules; match the set of association rules with rules in a knowledge base, and determine a matching result of the set of association rules; update information tuples in the knowledge base based on the interest degree, the matching result and the set of association rules; and determine a recommended course based on the information tuples in the knowledge base, and send the recommended course to the learning object; and the server comprises: a network application module and a network service module, the network application module is configured to provide an interface for managing the knowledge base, so as to update the knowledge base through the interface; and the network service module is configured to provide the updated knowledge base to the client.

[0008] According to a fourth aspect, an electronic device is provided, comprising: at least one processor; and a memory connected with the at least one processor in communication, wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method described in any implementation manner of the first aspect.

[0009] According to a fifth aspect, a non-transitory computer readable storage medium storing computer instructions is provided, the computer instructions are used to make a computer perform the method described in any implementation manner of the first aspect.

[0010] The embodiment of the present disclosure provides an information recommendation method and device. Firstly, a set of association rules is determined and sent based on a data set when a learning object completes an initial course. Secondly, the interest degree of a user to the set of association rules is obtained. Thirdly, the set of association rules is matched with rules in a knowledge base to determine a matching result of the set of association rules. Then, information tuples in the knowledge base are updated based on the interest degree, the matching result and the set of association rules. Finally, a recommended course is determined based on the information tuples in the knowledge base, and the recommended course is sent to the learning object. Thus, the set of association rules is matched with the rules in the knowledge base, and the information tuples in the knowledge base are updated, the dynamic and personalized course recommendation is realized through the information tuples, the reliability and accuracy of the recommended course are improved, and the experience of the learning object is improved. Meanwhile, the continuous updating of the knowledge base ensures the continuous optimization and adaptability of the recommendation system.

[0011] It should be understood that the content described in this part is not intended to identify key or important features of the embodiments of the present disclosure, nor to limit the scope of the present disclosure. Other features of the present disclosure will become apparent from the following description. BRIEF DESCRIPTION OF DRAWINGS

[0012] The accompanying drawings are used to better understand the present scheme, and do not constitute a limitation on the present disclosure. Among them:

[0013] Figure 1 is a flowchart of an embodiment of the information recommendation method according to the present disclosure;

[0014] Figure 2 is a structural schematic diagram of the main stages of the information recommendation method in the present disclosure;

[0015] Figure 3 is an architecture of a client and a server according to the present disclosure;

[0016] Figure 4 is a structural schematic diagram of an embodiment of the information recommendation device according to the present disclosure;

[0017] Figure 5 is a structural schematic diagram of an embodiment of the information recommendation system according to the present disclosure;

[0018] Figure 6 is a block diagram of an electronic device for implementing the information recommendation method of the embodiments of the present disclosure. DETAILED DESCRIPTION

[0019] Unless otherwise explicitly indicated, throughout the specification and claims, the term "comprise" or its variants such as "comprises" or "comprising" will be understood to encompass the stated element or components, without excluding the presence of other elements or components.

[0020] The technical solutions of the present disclosure are described below by means of specific examples. It should be understood that one or more steps mentioned in the present disclosure do not exclude other methods and steps before or after the combination steps, or other methods and steps can be inserted between the explicitly mentioned steps. It should also be understood that these examples are only used to illustrate the present disclosure and not to limit the scope of the present disclosure. Unless otherwise specified, the numbering of the steps of each method is only for the purpose of identifying the steps of each method, and not to limit the arrangement order of each method or to limit the scope of the implementation of the present disclosure. Changes or adjustments of the relative relationship can also be considered as the scope of the implementation of the present disclosure without substantial technical content changes.

[0021] The raw materials and instruments used in the examples are not specifically limited in source, and can be purchased on the market or prepared according to conventional methods well known to those skilled in the art.

[0022] Some current e-learning platforms have begun to use basic data mining and recommendation system technology. For example, some platforms recommend new learning resources based on the past learning records of students. However, these systems are usually basic and limited to simple content recommendation, and do not fully combine complex data mining and collaborative filtering algorithms for detailed analysis and optimization.

[0023] Many online education platforms collect students' learning behavior data such as course viewing time, click rate, etc., but these analyses mainly focus on descriptive statistical analysis, lacking in-depth pattern mining. Traditional analysis methods are usually static and cannot dynamically adjust teaching content.

[0024] In the field of e-commerce, recommendation systems have been widely used, mainly through collaborative filtering and association rule mining. For example, the recommendation system of a book selling platform will recommend related goods based on the user's historical purchase behavior. However, the main goal of commercial recommendation systems is to improve sales, which is different from the goal of improving teaching effectiveness in education systems, and there are differences in data processing and recommendation rules.

[0025] In view of the defects in the prior art, the present disclosure proposes an information recommendation method, which uses an intelligent recommendation system based on association rule mining and collaborative filtering technology to provide specific course improvement suggestions for course authors, and continuously optimizes teaching content to improve teaching quality and student learning effectiveness. Figure 1 The flow 100 of one embodiment of the information recommendation method according to the present disclosure is shown, and the information recommendation method includes the following steps:

[0026] Step 101, based on the data set when the learning object completes the initial course, determine and send the association rule set.

[0027] In this embodiment, the data set is the behavior interaction data of the learning object, and the learning object is a student or a student with a terminal. The behavior interaction data includes learning time, page clicks, course completion, grades, feedback, etc. The data set of the learning object completing the initial course is a multi-dimensional dynamic data set, which is mainly derived from the learning object interaction behavior and learning achievement data recorded in the learning management system (LMS). Specifically, it includes: the operation record of the learning object in the courseware of the initial course, such as the length of each unit learning, page click sequence, resource download frequency, test participation times, discussion area interaction, etc.; the scores of the learning object in each knowledge point test of the initial course, the quality of homework completion, the results of the stage examination, etc.; the learning object's score, comment on the initial course content, and questions or suggestions submitted during the learning process. The data set is indirectly related to the education object, which is a teacher or a teacher with a terminal. On the one hand, the behavior pattern in the data set directly reflects the influence of the course structure designed by the education object (such as knowledge point arrangement, resource distribution) on the learning path of the learning object; on the other hand, the teaching strategy score and feedback provided by the education object (such as the effectiveness evaluation of the teaching method for a certain type of knowledge point) will be integrated into the data set for optimizing the recommendation rules. Therefore, the data set is not only an objective record of the learning object's behavior, but also implies the subjective intention of the education object's teaching design.

[0028] In this embodiment, the association rule is a rule corresponding to an interesting relationship in the data set, and the association rule set includes at least one association rule. When the learning object completes the initial course, the execution subject on which the information recommendation method runs will analyze its learning behavior to generate a set of association rule sets, which reflect the learning mode, interest points and possible course requirements of the learning object.

[0029] In this embodiment, based on the data set when the learning object completes the initial course, the process of determining and sending the association rule set is one of the key steps in the information recommendation method. The core of this step is to analyze the behavior data of the learning object and mine potential association rules, thereby providing a basis for subsequent personalized recommendation.

[0030] Step 101 includes collecting initial data, preprocessing the initial data to obtain a data set, wherein the initial data is the behavior data of the learning object when completing the initial course, including but not limited to: course learning duration, course completion progress, test or homework score, interactive behavior such as clicking, browsing, and collecting, feedback on course content (such as rating, comment). Data preprocessing: clean, denoise and standardize the original data to ensure the quality and consistency of the data; use association rule mining algorithm to mine the behavior or course combination that often appears simultaneously in the learning process of the learning object from the preprocessed initial data, generate an initial rule set, for example: if the learning object has completed course A, they are likely to be interested in course B. If the learning object spends more time on a certain knowledge point, they may need additional auxiliary materials. The initial rule represents: each association rule can be represented as `X → Y`, where: `X` is the premise condition (such as having completed course A), and `Y` is the conclusion (such as recommending course B). According to the optimization goal, the initial rule set is screened and optimized to obtain an association rule set. The screening criteria for screening the association rule set from the initial rule set include: support, confidence, and lift. Among them, support is the frequency of the appearance of the association rule in the data set; confidence is the credibility of the rule, that is, the probability that the conclusion is established when the premise condition is established; lift is the relevance strength of the rule, which measures the independence between the premise condition and the conclusion; the optimization goal: retain rules with high support, confidence and lift, and eliminate low-quality or irrelevant rules.

[0031] In one example, the behavior data of the learning object is as follows: completed course A and spent a long time learning. After completing course A, frequently browsed content related to course B. The score for course A is high. Through association rule mining, the following rules may be generated: course A → course B (support: 0.8, confidence: 0.9, lift: 1.5); course A + high score → course C (support: 0.7, confidence: 0.85, lift: 1.4).

[0032] In this embodiment, the initial course can be an initial course that is initially generated to push to the learning object, and the initial course can also be a historical recommended course. The data set further includes: feedback of the learning object after completing the historical recommended course, expert verification results, and the association rule set can be obtained by mining the data set using an association rule algorithm. The association rule algorithm can be a Predictive Apriori algorithm.

[0033] Optionally, the set of association rules can be mined by a variety of association rule set mining algorithms, including: 1) Predictive Apriori algorithm for parameter-free association rule discovery; 2) IAS (Interestingness Analysis System) for subjective analysis, which classifies by comparing the unexpected rules with a previously defined knowledge database in the field of education. The educational objects with similar characteristics are found useful. The educational object characteristics are represented as a three-dimensional vector, which is related to his / her course with the following characteristics: subject (knowledge area, such as computer science or biology); level (course level, such as university, high school, elementary school or special education); difficulty (difficulty of the course, such as low or high). We use static classification to compare educational objects, so similar characteristics means a characteristic is exactly the same as other characteristics. A group of validation experts have voted on interest or usefulness. The implemented algorithm is particularly useful in collaborative recommendation systems, which can take advantage of the collaborative effect provided by the network to produce increasingly useful and accurate recommendations.

[0034] Step 102, obtaining the interest degree of the user to the set of association rules.

[0035] In this embodiment, the interest degree is the potential preference and acceptance degree of the user to each association rule in the set of association rules, and the interest degree can be measured by user feedback, click rate, learning time and other behavior data of the user to the recommended content. Optionally, the interest degree can also be measured by user feedback data and interaction data. Step 102 includes: collecting interaction behavior data of the user interacting with the set of association rules, for example: click rate: whether the user clicks the recommended course or content corresponding to the association rule. Learning time: the time spent by the user on the recommended course corresponding to the association rule. Completion rate: whether the user completes the recommended course corresponding to the association rule. Collection, sharing, commenting and other behaviors. Explicit feedback data of the user to the data corresponding to the set of association rules is adopted, and the explicit feedback data includes: scoring, liking / disliking and questionnaire survey, for example: scoring: the user's score (such as 1-5 stars) on the recommended content corresponding to the set of association rules. Liking / disliking: the user's expression of interest in the recommended content corresponding to the set of association rules. Questionnaire survey: directly obtaining the interest degree of the user to the recommended content corresponding to the set of association rules through the questionnaire.

[0036] Based on user behavior data, implicit interest degree is calculated through statistical or machine learning methods. For example: the higher the click rate, the higher the interest degree. The longer the learning time, the higher the interest degree. The higher the completion rate, the higher the interest degree. Based on user feedback data, direct quantitative display interest degree. For example: the higher the user rating, the higher the interest degree. The more the user likes, the higher the interest degree. Combine implicit and explicit interest degrees, and calculate the interest degree of the association rule set by weighting or other methods.

[0037] In this embodiment, the system can update the user's interest degree in real time as the user continues to interact with the recommended content. For example, if the user shows a high click rate and learning time for the newly recommended course, increase its interest degree. Long-term update: Recalculate the user's interest degree regularly to ensure that the recommendation system can adapt to changes in user interest.

[0038] In one example, assume the association rule set is: if the user has completed "Python Programming Basics", then recommend "Data Science Practice". If the user has completed "Data Analysis for Beginners", then recommend "Machine Learning Basics". The user clicked on the "Data Science Practice" course and learned for 30 minutes. The user rated the "Machine Learning Basics" course with 4 stars. Based on these data, calculate the user's interest degree for the two recommended contents: interest degree for "Data Science Practice": based on click rate and learning time, interest degree is 0.8. Interest degree for "Machine Learning Basics": based on rating, interest degree is 0.9.

[0039] Step 103, match the association rule set with the rules in the knowledge base to determine the matching result of the association rule set.

[0040] In this embodiment, the generated set of association rules is matched with the existing rules in the knowledge base to determine the matching results. The matching results include: matching with the rules in the knowledge base, partially matching with the rules in the knowledge base, and not matching with any rule in the knowledge base. The matching results can help the system determine whether the current rule set is consistent with the rules in the existing knowledge base or whether the knowledge base needs to be updated. In this embodiment, a rule matching algorithm can be used to match the set of association rules with the rules in the knowledge base. Common matching processes include: exact matching, which determines whether the premise and conclusion of the association rule are completely consistent with a certain rule in the knowledge base. Partial matching, which determines whether the premise or conclusion of the association rule is partially consistent with a rule in the knowledge base (for example, the premise contains the same course or knowledge point). Similarity matching, which calculates the similarity of the association rule and the rule in the knowledge base (such as based on text similarity or semantic similarity), and sets a threshold to determine whether it matches. Matching result classification: according to the matching results, the matching results of the set of association rules are classified into the following categories: complete matching, the association rule is completely consistent with a certain rule in the knowledge base; partial matching, the association rule is partially consistent with a certain rule in the knowledge base; new rule, the association rule does not match any rule in the knowledge base.

[0041] In this embodiment, for complete matching, if the association rule completely matches a certain rule in the knowledge base, the rule in the knowledge base can be directly used for recommendation, and the knowledge base does not need to be updated. For partial matching, if the association rule partially matches a certain rule in the knowledge base, the rule in the knowledge base can be updated or expanded in combination with the user interest degree and the matching result. For example: improving the confidence of the matching rule. Extending the premise or conclusion of the rule. For new rules, if the association rule is completely new, it can be added to the knowledge base as a new information tuple.

[0042] Optionally, the step 103 comprises: matching the set of association rules with the rules in the knowledge base using a subjective analysis algorithm to determine the matching results of the set of association rules. The subjective analysis algorithm is a rule matching method based on domain knowledge, expert experience or manual judgment. Unlike traditional automated matching algorithms, the subjective analysis algorithm places more emphasis on the semantic similarity between rules and the degree of fit of business logic. When the rule matching needs to consider context, semantics or domain knowledge, the subjective analysis algorithm can make up for the shortcomings of automated algorithms. The set of association rules and the rules in the knowledge base are converted into a unified representation form for matching. For example: the premise conditions and conclusions of the rules are represented as structured data. The semantic information of the rules is represented as natural language text; the semantic similarity between rules is calculated using natural language processing (NLP) techniques. For example: the semantic similarity of the premise conditions and conclusions is calculated using a word vector model (such as Word2Vec, BERT). A similarity threshold is set to determine whether the rules match. The context information of the rules (such as course category, learner characteristics) is combined for matching. For example: if the premise conditions of two rules belong to the same course category, they are considered to be partially matched.

[0043] In this embodiment, in the subjective analysis stage, the execution subject on which the information recommendation method runs matches the newly discovered association rules with the existing rules in the knowledge base through the IAS algorithm (Interestingness Analysis System). During the matching process, the system calculates the values of four subjective indicators (i.e. conform_ij, unexpConseq_ij, unexpCond_ij, bsUnexp_ij) for each new rule. These values are between 0 and 1, representing the matching degree of the new rule with the existing rules in the knowledge base in terms of condition conformity, result unexpectedness, condition unexpectedness or bilateral unexpectedness. For example, if the condition part of a new rule is completely identical to a rule in the knowledge base, but the result part is different, its unexpConseq_ij value may be close to 1, and it is classified as an "unexpected result rule".

[0044] Step 104: updating the information tuples in the knowledge base based on the interest degree, the matching results and the set of association rules.

[0045] In this embodiment, based on the user's interest degree, the matching results and the set of association rules, the system dynamically updates the information tuples in the knowledge base. This step ensures that the knowledge base can be continuously optimized as the user's learning behavior and interests change.

[0046] In this embodiment, the information tuple is a storage unit in the knowledge base, and the information tuple includes rules, problems, recommended information, and relevance. The problem in the information tuple refers to a specific pain point in a teaching scenario or course design corresponding to the rule discovered in the association rule mining stage. For example, it can be discovered through association rule analysis that "after students complete unit A, the passing rate of unit B significantly decreases", and the teaching problem (such as insufficient knowledge connection between units A and B) reflected by this rule is abstracted as "problem".

[0047] In this embodiment, the recommended information in the information tuple is a specific improvement suggestion generated based on the rules and the collaborative filtering result. For example, for the above "problem", the recommendation can be "add transitional exercises in unit A or adjust the teaching order of unit B". The generation of the recommended information can be combined with two sources: first, the possible improvement direction is directly deduced through the IF-THEN rule. Second, combined with the historical scores and feedback of the education object and the experts (such as verified strategies for similar problems), an effective solution recognized by peers is recommended. The lack of the "recommendation" field in the information tuple will cause the rule to only stay at the phenomenon description level and cannot be converted into an operable course optimization measure.

[0048] In this embodiment, the relevance in the information tuple is used to quantify the matching degree of the recommendation and the problem, and its judgment depends on the multi-dimensional score in the collaborative filtering module. Specifically: the education object votes for the practicability of the recommendation through the client to form an initial relevance score; the expert team verifies and supplements the score of the recommendation to ensure its scientificity and universality; and finally the scores of the education object and the experts are integrated through a formula to reflect the comprehensive value of the recommended information. If the "relevance" field is lacking, the system will not be able to distinguish the quality priority of the recommendation, and the education object will need to spend additional effort to filter effective suggestions, reducing the efficiency of the system. Through step 104, the knowledge base can be dynamically optimized to better reflect the changes in the learning behavior and interest of the user, thereby providing more accurate recommendations.

[0049] In this embodiment, step 104 includes: in response to the matching result being that an association rule in the association rule set completely matches the rule in the information tuple of the knowledge base, updating the recommended information and the relevance in the information tuple according to the user interest degree and the matching result.

[0050] In this embodiment, step 104 can further include: in response to the matching result being that the association rule is completely new, generating a new information tuple through the association rule, and adding the new information tuple to the knowledge base.

[0051] Optionally, step 104 includes: selecting an association rule with a top interest degree from the association rule set, obtaining multiple types of matching results based on the association rule after rule matching using the subjective analysis algorithm, and updating the information tuples in the knowledge base through classification of the matching results. Specifically:

[0052] Conformity rule (high conform_ij value): If the new rule is highly consistent with the existing rules in the knowledge base, it is marked as a "known effective rule" and can directly update the relevance score of the corresponding tuple in the knowledge base (such as increasing the voting weight), without adding new tuples.

[0053] Unexpected result / condition rule (high unexpConseq_ij or unexpCond_ij value): This type of rule reflects new causal relationships or potential problems. For example, students do not master unit B as expected after completing unit A, but the system finds that the completion rate of unit C has significantly improved. At this time, the system will convert this type of rule into a new rule tuple, including the association rule itself, the corresponding teaching problem (such as "the association between unit A and unit C is not fully utilized"), recommended information (such as "add guided content for unit C after unit A"), and assign an initial relevance score based on expert or learner voting.

[0054] Bilateral unexpected rule (high bsUnexp_ij value): This type of rule may reveal new learning patterns or design flaws and needs to be further verified by experts. The system will mark it as a "high-priority rule to be reviewed" and generate a temporary tuple to be stored in the knowledge base, which will be converted into a formal tuple or deleted after expert confirmation.

[0055] The specific logic for revising the knowledge base is that after determining the rule type through the matching results, the system automatically associates the new rule with the problem-recommendation framework in the knowledge base, dynamically adjusts the priority of the tuple (for example, unexpected rules are assigned a higher initial weight), and updates the relevance of the tuple based on voting and prediction accuracy (PA algorithm). This process ensures the continuous optimization of the knowledge base and converts subjective analysis results into actionable course improvement recommendations.

[0056] Step 105, based on the information tuples in the knowledge base, determine the recommended course and send it to the learning object.

[0057] In this embodiment, the information recommendation method generates personalized course recommendations based on the information tuples in the knowledge base and sends them to the learning object. The recommended courses will be more tailored to the user's learning needs and interests, improving learning effectiveness and user experience.

[0058] In the embodiment, the determining the recommended course based on the information tuples in the knowledge base comprises: extracting recommended information of all information tuples in the knowledge base, and taking a course with the most frequency or times in the recommended information of all information tuples as the recommended course.

[0059] Optionally, after the recommended course is sent to the learning object, subsequent learning behavior data of the learning object is collected, and the learning behavior data of the learning object is again input into an execution subject in which the information recommendation method runs, and the knowledge base is updated by again extracting the association rule set, to form a closed loop iteration.

[0060] The information recommendation method provided by the embodiment of the present disclosure first determines and sends the association rule set based on the data set when the learning object completes the initial course, secondly, obtains the interest degree of the user to the association rule set, thirdly, matches the association rule set with the rules in the knowledge base to determine the matching result of the association rule set, then, updates the information tuples in the knowledge base based on the interest degree, the matching result and the association rule set, and finally, determines the recommended course based on the information tuples in the knowledge base and sends the recommended course to the learning object. Thus, the association rule set is matched with the rules in the knowledge base, and the information tuples in the knowledge base are updated, the dynamic and personalized course recommendation is realized through the information tuples, the reliability and accuracy of the recommended course are improved, and the experience of the learning object is improved. Meanwhile, the continuous updating of the knowledge base also ensures the continuous optimization and adaptability of the recommendation system.

[0061] In one specific example, as shown in FIG. 1, it is the main stage of the information recommendation method of the present disclosure, and the main stage comprises: a first stage to a fourth stage. In the first stage to the fourth stage, each stage plays its own role, and finally the recommended course is obtained. Figure 2

[0062] ​The first stage is association rule set mining: the goal of this stage is to find association rule sets on the data set generated when the learning object finishes the course. Once the data is preprocessed, the output of this stage, i.e., the determined association rule set, then enters the second stage for subjective analysis. This stage uses a subjective rule evaluation metric to determine the interest degree of the rules found by the association rule set mining. It also applies the IAS algorithm to compare these rules with the rules in the knowledge base to classify them as expected or unexpected. Among them, the subjective rule evaluation metric refers to the evaluation method for the value judgment of the rules generated by the association rule mining through the professional knowledge or experience of the user (such as the education object, the education expert). Its core lies in combining the subjective cognition of the user, rather than relying only on objective statistical indicators (such as support and confidence). Specifically, the interest degree of the rule represents the degree of attention and practicality evaluation of the rule by the courseware author or the education expert. For example, a rule may show that "learning objects who complete unit A have an 80% probability of mastering unit B", if the rule is consistent with the teaching experience of the education object, it is classified as "consistent rule", and the interest degree is high; if the rule reveals an unexpected association of the education object (such as "learning objects who frequently use a certain interactive tool have significantly improved performance"), it may be marked as "unexpected rule", and the interest degree is also high, because it provides a new optimization direction.

[0063] The "expected" rule refers to the new rule found by the association rule mining being highly consistent with the existing rules in the knowledge base or the priori knowledge in the education field, i.e., the content, logical relationship or application scenario of these rules conform to the existing experience of the education object and the education expert or the verified teaching strategies. For example, if there is a rule in the knowledge base that "after learning objects complete unit A, the passing rate of unit B is significantly improved", and the newly mined rule "after learning objects complete unit A, the completion time of unit B is shortened" has strong association in the condition and result, it can be classified as "expected".

[0064] The "unexpected" rule refers to the rule that is inconsistent with the knowledge base. The creation of the knowledge base is a "one-time initialization and continuous optimization" process. Each time the system runs, it is based on the existing knowledge base, combined with new data and feedback for incremental update, rather than rebuilding from scratch. This design not only guarantees the stability of the knowledge base and the reuse of historical experience, but also realizes dynamic evolution through the synergy mechanism, and finally forms an intelligent rule base with self-improving ability.

[0065] The third stage is the knowledge base updating stage: this stage combines collaborative filtering techniques and knowledge-based techniques to update the rule base. The knowledge base stores information tuples (rule-question-recommendation information-relevance) and is classified according to specific course profiles. To avoid the cold start problem of the collaborative filtering system, the first information tuples of the repository are proposed by experts and voted by other experts. On the other hand, the education objects can find new information tuples, which can be inserted into the repository after being verified by expert objects and voted by other tuples.

[0066] The fourth stage is the recommendation stage, in which the expected rules found in the second stage are combined with the more intuitive information tuples mentioned in the third stage and then used in the fourth stage to show the education objects, who are not good at data mining in most cases, some possible solutions to the problems detected in the course. The education objects analyze the recommended content and determine whether it is relevant, and finally determine the recommended course.

[0067] Among them, the content of the recommendation information about the fourth stage mainly includes: for example, "the interaction rate of unit C is lower than the average level, which may affect the mastery of knowledge points". Based on the association rules and the recommendation field in the tuple, such as "it is recommended to add a simulation experiment module in unit C (from rule R001)" "reference to the interactive question and answer strategy adopted by other education objects in similar courses (collaborative filtering recommendation)". The support, confidence, education object / expert voting ratio of the rule are displayed in a visual form, such as "this suggestion has been adopted by 85% of education objects in similar courses" "expert rating 4.8 / 5.0". Provide modification options that can be directly embedded in courseware (such as inserting specific resource links, adjusting chapter order), and support one-key application or further customization.

[0068] In some embodiments of the present disclosure, the above information recommendation method further comprises: obtaining a data set of the learning object completing the recommended course; determining and sending a new association rule set based on the data set of the learning object completing the recommended course; generating a new recommended course based on the new association rule set, the interest degree of the user to the new association rule set, and the matching result of the new association rule set in the knowledge base, and sending the new recommended course to the learning object.

[0069] In this embodiment, the data set of the learning object completing the recommended course includes: the operation record of the learning object in the courseware of the recommended course, the score, the homework completion quality, the stage examination result, etc. of the learning object in each knowledge point test of the recommended course, the score, the comment and the question or suggestion submitted in the learning process of the learning object to the content of the recommended course.

[0070] In the embodiment, the method for determining and sending the new association rule set based on the data set of the learning object completing the recommended course comprises: selecting an evaluation result of the learning object from the data set of the learning object completing the recommended course, such as a score, a comment, a question or a suggestion, and updating the association rule set based on the evaluation result to obtain the new association rule set.

[0071] In the embodiment, the method for obtaining the interest degree of the user for the new association rule set can refer to the method for obtaining the interest degree of the user for the association rule set, and the method for obtaining the matching result of the new association rule set in the knowledge base can refer to the method for obtaining the matching result of the association rule set in the knowledge base.

[0072] The information recommendation method provided in the embodiment comprises: obtaining a data set of a learning object completing a recommended course; determining and sending a new association rule set based on the data set of the learning object completing the recommended course; generating a new recommended course based on the new association rule set, the interest degree of the user for the new association rule set and the matching result of the new association rule set in the knowledge base, and sending the new recommended course to the learning object, thereby ensuring the iterative optimization of the information in the knowledge base and the recommended information and improving the accuracy and reliability of the new recommended course.

[0073] In some embodiments of the present disclosure, the method for determining and sending the association rule set based on the data set of the learning object completing the initial course comprises: preprocessing the data set of the learning object completing the initial course to obtain preprocessed data; performing rule analysis on the preprocessed data and a search area selected by the education object by using an association analysis algorithm to obtain the association rule set; and sending the association rule set to the education object.

[0074] In the embodiment, the preprocessing is data cleaning, normalization processing and the like of the data set, and the preprocessing can make the preprocessed data suitable for the association rule analysis. The search area is a data range in the preprocessed data, such as a data specification and a data attribute, and the search area selected by the education object can simplify the positioning area.

[0075] In the embodiment, the method for determining and sending the association rule set based on the data set of the learning object completing the initial course comprises the following specific implementation steps and methods: step 1: data preprocessing, the data set of the learning object completing the initial course is cleaned and arranged and the like to make it suitable for the subsequent association rule analysis; step 2: association rule analysis, the data and attributes in the search area selected by the education object in the preprocessed data are analyzed by using an association analysis algorithm to mine the association rules between the courses. The generated association rule set is sent to the education object (such as an education object, an administrator or a recommendation system).

[0076] The method for determining and sending the association rule set provided by the present disclosure includes preprocessing a data set of a learning object when the learning object finishes an initial course to obtain preprocessed data; performing rule analysis on the preprocessed data and a search area selected by an education object by using an association analysis algorithm to obtain an association rule set; and sending the association rule set to the education object. The accuracy and reliability of the association rule set are improved.

[0077] Optionally, the method for determining and sending the association rule set based on the data set of the learning object when the learning object finishes the initial course includes preprocessing the data set of the learning object when the learning object finishes the initial course to obtain preprocessed data; performing rule analysis on the preprocessed data by using an association analysis algorithm to obtain an association rule set; and sending the association rule set to the education object.

[0078] In some optional implementations of the present disclosure, the method for obtaining the interest degree of the user for the association rule set includes obtaining feedback data of the user for each association rule in the association rule set; determining a first voting proportion of the education object for each association rule and a second voting proportion of an expert object for each association rule based on the feedback data; and determining the interest degree of each association rule in the association rule set based on the first voting proportion and the second voting proportion.

[0079] In the optional implementation, the user includes the education object and the expert object, the education object is an object that outputs a recommended course to the learning object, such as an education object or an education object with an intelligent terminal, the education object evaluates the association rules in the association rule set after obtaining the association rule set and gives feedback, the expert object is an object that comprehensively evaluates the education object and the learning object, such as an expert or an expert with an intelligent terminal, the expert object evaluates the association rules in the association rule set after obtaining the association rule set and gives feedback, and the feedback of the education object and the feedback of the expert object are used as the feedback data of the user.

[0080] In the optional implementation, the feedback data of the user for each association rule in the association rule set includes providing a feedback interface or a tool for the user to evaluate the association rules. For example, providing options of “agree”, “disagree”, or “neutral” or providing a scoring function (such as 15 stars).

[0081] In the optional implementation, the first votes of the education objects and the second votes of the expert objects in the feedback data are distinguished, the number of the first votes and the number of the second votes are taken as the total number of votes, the first vote proportion is equal to the number of the first votes of each association rule divided by the total number of votes, and the second vote proportion is equal to the number of the second votes of each association rule divided by the total number of votes. The first vote proportion and the second vote proportion of each association rule are weighted and summed to obtain the interest degree of each association rule, where the weight of the first vote proportion and the weight of the second vote proportion can be set based on requirements, for example, the weight of the second vote proportion is greater than the weight of the first vote proportion.

[0082] Specifically, the weighted sum of the first vote proportion and the second vote proportion of each association rule to obtain the interest degree of each association rule includes: the weighted sum of the first vote proportion and the second vote proportion to obtain the interest degree. For example: interest degree = w1 x first vote proportion + w2 x second vote proportion; where w1 and w2 are weight parameters, and w1 + w2 = 1. The weight can be adjusted according to business requirements. If more attention is paid to expert opinions, w2 can be set to 0.7 and w1 to 0.3.

[0083] The method for determining the interest degree of each association rule in the association rule set provided by the optional implementation includes: obtaining feedback data of users on each association rule in the association rule set; determining a first vote proportion of the education objects on each association rule and a second vote proportion of the expert objects on each association rule based on the feedback data; and determining the interest degree of each association rule in the association rule set based on the first vote proportion and the second vote proportion, thereby improving the reliability and accuracy of the interest degree.

[0084] Optionally, the obtaining of the interest degree of the user on the association rule set includes: obtaining feedback data of the user on each association rule in the association rule set; determining a set of association rules approved by the user in the association rule set based on semantic information of the feedback data, and determining an approval degree value of the association rule based on the semantic information, taking the approval degree value of each association rule as the interest degree of each association rule in the association rule set.

[0085] In some optional implementations of the present disclosure, the determination of the interest degree of each association rule in the association rule set based on the first vote proportion and the second vote proportion includes: determining a first weighting coefficient of the education objects and a second weighting coefficient of the expert objects; obtaining a weighted metric value based on the first vote proportion, the second vote proportion, the first weighting coefficient, and the second weighting coefficient; performing accuracy prediction on the voting results of the education objects on each association rule to obtain an accuracy result of each association rule; and obtaining the interest degree of each association rule based on the weighted metric value and the accuracy result.

[0086] In the optional implementation, For different data sources Different educational objects, for The discovered association rule set, ;set up for According to the weight definition, Rules in Votes can be used to Assign weight .

[0087] In fact, the education subjects are more inclined to apply those association rules that have received more support or more votes, and the first vote share is shown in formula (1).

[0088] (1)

[0089] In formula (1), , is Middle Pair Rules The number of education subjects who voted.

[0090] By applying the same reasoning to expert object voting, the second voting proportion is shown in Equation (2):

[0091] (2)

[0092] In formula (2), , is Middle Pair Rules The number of experts who voted.

[0093] Therefore, the rules The weight of can be expressed as a weighted measure of the votes of education objects and expert objects, that is:

[0094] (3)

[0095] In formula (3), and are the first weighting coefficient and the second weighting coefficient representing the education object and the expert object respectively.

[0096] Once the weight of each rule is calculated, an interest metric can be designed, also known as weighted accuracy. According to the PA algorithm, the prediction accuracy of the rule can be calculated for the rule. Defining Interest As shown in formula (4):

[0097] (4)

[0098] wherein is a weighted measure value calculated according to formula (3), is the prediction accuracy result returned by the PA algorithm for each rule of the voting support rule set.

[0099] In this embodiment, the PA algorithm (Predictive Apriori algorithm) provides a prediction accuracy result for each rule. The accuracy result is calculated by the PA algorithm through the Bayesian method, reflecting the prediction ability of the rule for future data, and is used to comprehensively measure the effectiveness and practicality of the rule. By taking the prediction accuracy as an important basis for rule sorting and recommendation, the priority and credibility of the system recommendation to the courseware author are ultimately affected.

[0100] In some optional implementations of the present disclosure, the above updating of the information tuples in the knowledge base based on the interest degree, the matching result and the set of association rules comprises: screening a set of to-be-transformed rules from the set of association rules based on the interest degree; screening matching results related to the set of to-be-transformed rules from the matching result; transforming the set of to-be-transformed rules into information tuples based on the matching result of the set of to-be-transformed rules; and updating the information tuples in the knowledge base by using the transformed information tuples.

[0101] In this optional implementation, the association rule with the highest interest degree is determined, and the selected association rule is taken as the set of to-be-transformed rules, and the set of to-be-transformed rules is screened from the set of association rules. The above screening of matching results related to the set of to-be-transformed rules from the matching result comprises: based on the association rules in the set of to-be-transformed rules, taking the matching results related to the association rules as the matching results related to the set of to-be-transformed rules.

[0102] In this optional implementation, the matching results related to the set of to-be-transformed rules are screened from the matching result according to the set of to-be-transformed rules. For example, if the set of to-be-transformed rules contains rule A, the matching result of rule A is screened. The information tuples are usually represented in the form of structured data, such as rules, questions, recommended information and correlations. Compliance: if the matching result is “compliance”, the information tuples in the knowledge base are directly used without transformation. Unexpected result: if the matching result is “unexpected result”, a new information tuple is generated, the premise condition is retained, and the conclusion is updated. Unexpected condition: if the matching result is “unexpected condition”, a new information tuple is generated, the conclusion is retained, and the premise condition is updated. Double unexpected: if the matching result is “double unexpected”, a completely new information tuple is generated.

[0103] ​In the optional implementation, the converting the to-be-converted rule set into information tuples based on the matching result of the to-be-converted rule set comprises: generating information related to the rule, such as a teaching question, recommended information, and a correlation score, according to the rule in the to-be-converted rule set. For example, the rule is "unit A -> unit B", the corresponding teaching question is "low unit B completion rate", the recommended information is "suggest adding exercises after unit A", and the correlation score is generated based on voting and prediction accuracy.

[0104] In the optional implementation, the updating the information tuples in the knowledge base by using the converted information tuples comprises: storing the converted information tuples in the knowledge base according to course characteristics (theme, difficulty, and level) so that the information tuples in the knowledge base support collaborative filtering matching.

[0105] Optionally, the updating the information tuples in the knowledge base by using the converted information tuples further comprises: adding the converted information tuples to the knowledge base and removing duplicate information tuples. The converted information tuples are inserted into the knowledge base. It is checked whether there is a rule (such as a premise condition and a conclusion) that is completely the same as the newly added information tuple in the knowledge base. If there is, the duplicate information tuple is removed.

[0106] The method for updating the information tuples in the knowledge base provided by the present disclosure can dynamically update the information tuples in the knowledge base based on interest degree, matching result, and association rule set, thereby providing more accurate support for subsequent personalized recommendation.

[0107] In some optional implementations of the present disclosure, the updating the information tuples in the knowledge base by using the converted information tuples comprises: adding the converted information tuples to the knowledge base and removing duplicate information tuples in the knowledge base; collecting prediction accuracy of each information tuple in the knowledge base; determining the correlation of each information tuple; determining the priority of each information tuple based on the prediction accuracy and the correlation; and sorting the information tuples in the knowledge base based on the priority of each information tuple.

[0108] In the optional implementation, the prediction accuracy refers to the recommendation accuracy of the rule in actual application, which is usually calculated by the following formula: PA = number of correct recommendations / total number of recommendations, wherein the number of correct recommendations and the total number of recommendations of each rule are counted through historical recommendation data or experimental data.

[0109] Optionally, the prediction accuracy of each information tuple in the knowledge base is determined by using a PA (Apriori) algorithm, wherein the PA algorithm provides an accurate prediction accuracy for each rule. The prediction accuracy is calculated by the PA algorithm through a Bayesian method, and reflects the prediction ability of the rule for future data. The prediction accuracy result generated by the PA algorithm is integrated into the comprehensive evaluation of the rule, and finally affects the priority and credibility of the recommendation of the system to the courseware author.

[0110] Optionally, the prediction accuracy can also be combined with the voting proportion of the education object and the expert object to form a weighted accuracy (WAcc), which is used to comprehensively measure the effectiveness and practicality of the rule.

[0111] In the optional implementation, the relevance of each information tuple in the knowledge base is determined to evaluate the importance of the rule. The relevance refers to the degree of fit of the rule to the current learning object or business requirement, and the relevance score of each rule can be calculated based on the usage frequency (such as the number of recommendations) of the rule, based on the interest degree (such as user feedback data) of the rule, and based on the semantic relevance (such as the matching degree with the current learning object) of the rule.

[0112] In the optional implementation, the priority of each information tuple is determined based on the prediction accuracy and the relevance, which includes: weighted sum of the prediction accuracy and the relevance to obtain a priority score. Normalization: normalize the priority score to 0 to 1, which is convenient for subsequent sorting. Calculate the priority of each information tuple, and sort the information tuples in the knowledge base. The information tuples are sorted according to the priority score from high to low. Update the sorted information tuples in the knowledge base to ensure that the rules with high priority can be used preferentially.

[0113] The method for updating the information tuples in the knowledge base provided by the optional implementation can update the information tuples in the knowledge base by using the transformed information tuples, and sort the rules based on the prediction accuracy and the relevance, thereby providing more accurate support for subsequent personalized recommendation.

[0114] In some optional implementations of the present disclosure, the information tuples in the knowledge base are constructed by the following steps: obtaining cold-start rules of expert objects; determining expert tuples based on the cold-start rules; obtaining education tuples sent by education objects; sending the education tuples to the expert objects; in response to receiving a verification pass result of the expert objects on the education tuples, regarding both the expert tuples and the education tuples as information tuples in the knowledge base.

[0115] In this optional implementation, the cold start rule is a rule that the expert object sets in advance for the knowledge base at the cold start. The determining of the expert tuple based on the cold start rule includes: determining a question in a teaching scenario or course design based on the cold start rule; determining recommendation information based on the question, and sending the question and the recommendation information; obtaining the relevance of the question and the recommendation information; and taking the cold start rule, the question, the recommendation information, and the relevance as the expert tuple.

[0116] In this optional implementation, the education tuple is a tuple that the education object sets for the cold start of the knowledge base. The structure of the education tuple can be: rule, question, recommendation information, and relevance.

[0117] In this optional implementation, in order to avoid the cold start problem of the collaborative filtering system, the expert proposes the first tuple of the repository, and votes on the tuples proposed by other experts. On the other hand, the teacher can find new tuples, which must be verified by experts before being inserted into the repository, and votes on other tuples, thereby improving the reliability of the cold start of the knowledge base.

[0118] In some optional implementations of the present disclosure, the information tuple is constructed by the following steps: determining an association rule; determining a question in a teaching scenario or course design based on the association rule; determining recommendation information based on the question, and sending the question and the recommendation information; obtaining the relevance of the question and the recommendation information; and taking the association rule, the question, the recommendation information, and the relevance as the information tuple.

[0119] In this optional implementation, the association rule can be extracted from the behavior data of the learning object by an association rule mining algorithm (such as Apriori, FPGrowth). The specific steps of extracting the association rule include: collecting the behavior data of the learning object (such as course completion records, test scores, learning duration, etc.). An association rule set is generated using an association rule mining algorithm. Each rule is usually represented in the form of “IF THEN”, for example: if the user has learned “Python Programming Basics”, then recommend “Data Science Practice”.

[0120] In this optional implementation, the determining of the question in the teaching scenario or course design based on the association rule includes:

[0121] By the premise and conclusion of the association rule, potential teaching problems are identified, for example: if the user does not complete “Data Science Practice” after learning “Python Programming Basics”, there may be a course connection problem; the identified questions are classified (such as course design problems, learning path problems, knowledge point coverage problems, etc.).

[0122] In the optional implementation, the determining the recommendation information based on the question includes: generating the recommendation information according to the type of the question and the association rule. For example, if there is a course connection question, recommending “Data Science Practice” as a subsequent course; if there is a knowledge point coverage question, recommending supplementary learning materials. The recommendation information is converted into structured data (such as JSON, XML).

[0123] In the optional implementation, the correlation refers to the degree of fit between the recommendation information and the question, and the correlation score between the recommendation information and the question can be calculated based on user feedback data (such as scores, click rates) and actual effects of the recommendation information (such as completion rates of learning objects, test scores).

[0124] The method for constructing the information tuple provided by the present disclosure can construct the information tuple and use it for updating the knowledge base and personalized recommendation, thereby improving the teaching effect and learning experience.

[0125] In some optional implementations of the present disclosure, the determining the recommendation course based on the information tuple in the knowledge base and sending the recommendation course to the learning object includes: determining the ranking of each information tuple based on the information tuple in the knowledge base; determining the information tuple to be recommended based on the ranking; extracting the recommendation information in the information tuple to be recommended and sending the recommendation information to the education object; receiving the recommendation course sent by the education object and sending the recommendation course to the learning object.

[0126] In the optional implementation, the determining the recommendation course based on the information tuple in the knowledge base includes: ranking the information tuples in the knowledge base according to the priority, interest degree, and confidence of the information tuples. The priority is calculated based on the prediction accuracy and correlation, the interest degree is obtained based on the user feedback data, and the confidence is obtained based on the confidence of the rule. When ranking the information tuples, the information tuples can be ranked from high to low according to the ranking indexes. For example, the information tuples with high priority are ranked in the front.

[0127] In the optional implementation, the determining the information tuple to be recommended based on the ranking includes: setting a filtering condition according to the business requirement, and the filtering condition includes: selecting the first N information tuples with the highest priority. Selecting the information tuples with an interest degree greater than a certain threshold.

[0128] In the optional implementation, the information tuple to be recommended includes multiple recommendation information, and sending the recommendation information to the education object can make the education object recommend the recommendation course most related to the multiple recommendation information, thereby facilitating the sending of the recommendation course to the learning object.

[0129] The method for determining recommended courses provided by this optional implementation can determine recommended courses based on information tuples in the knowledge base and send them to the learning object, thereby improving the learning experience and effect. The information recommendation method provided by this disclosure is based on a client-server architecture. Figure 3 As shown, there are multiple clients that apply association rule mining algorithms locally on the data of online courses used by learning objects. The server application consists of two modules. The first is a web application server (for installing Figure 3 The second module is a web service that allows the server to share updated KBs with clients in PMML format. PMML (Predictive Model Markup Language) is an XML-based language that makes it possible to define and share predictive models between different applications, establishing a vendor-independent way of defining these models and thus avoiding proprietary applications and compatibility issues. Therefore, once an updated version of the KB has been downloaded from the server, the client can apply the mining algorithm offline. The client application is part of the iterative methodology used by educators to develop courses. It is able to detect potential issues in the design and content of e-learning courses by adding feedback or maintenance phases to the courses.

[0130] As shown in Figure 3, there are several stages in this methodology: 1) Initial construction of the course, e.g. Figure 3 1) Educational object J creates or modifies the course; 2) Learning object X completes the course, i.e. Figure 3 During the course completion, usage information is transparently compiled and stored in the knowledge base ( Figure 3 The data used in the process is the data in the knowledge base); 3) the continuous improvement phase, which is carried out in parallel with the client application. The last phase contains the core of the rule mining algorithm used. The algorithm and the knowledge base classify the found rules as expected (if consistent with the knowledge base) or unexpected (if inconsistent), corresponding to Figure 3 "Recommendations" and "surprises" in [1]. If education subject J applies a recommendation to a course, they implicitly vote on its usefulness in the server knowledge base. Unexpected tuples are ranked according to the IAS algorithm, and education subjects can mark any interesting tuples. Expert subject Z then analyzes these unexpected "interesting" or "valid" information tuples and may choose to include them in the knowledge base.

[0131] Further references Figure 4 As an implementation of the methods shown in the above figures, the present disclosure provides an embodiment of an information recommendation device. Figure 1The method embodiments shown correspond to the device, which can be specifically applied to various electronic devices.

[0132] As shown in Figure 4 The information recommendation device 400 provided by the embodiment includes a determination unit 401, an acquisition unit 402, a matching unit 403, an updating unit 404, and a recommendation unit 405. The determination unit 401 can be configured to determine and send a set of association rules based on a data set when a learning object completes an initial course. The acquisition unit 402 can be configured to acquire an interest degree of a user for the set of association rules. The matching unit 403 can be configured to match the set of association rules with rules in a knowledge base, and determine a matching result of the set of association rules. The updating unit 404 can be configured to update information tuples in the knowledge base based on the interest degree, the matching result, and the set of association rules. The recommendation unit 405 can be configured to determine a recommended course based on the information tuples in the knowledge base, and send the recommended course to the learning object.

[0133] In the embodiment, the specific processing of the determination unit 401, the acquisition unit 402, the matching unit 403, the updating unit 404, and the recommendation unit 405 in the information recommendation device 400 and the technical effects brought by the specific processing can be referred to the corresponding descriptions of the steps 101, 102, 103, 104, and 105 respectively. Figure 1 The related descriptions of the steps 101, 102, 103, 104, and 105 in the corresponding embodiments will not be repeated here.

[0134] In some embodiments of the disclosure, the determination unit 401 is further configured to: preprocess the data set when the learning object completes the initial course to obtain preprocessed data; perform rule analysis on the preprocessed data and a search area selected by the education object by using an association analysis algorithm to obtain the set of association rules; and send the set of association rules to the education object.

[0135] In some embodiments of the disclosure, the acquisition unit 402 is further configured to: acquire feedback data of the user for each association rule in the set of association rules; determine a first voting proportion of the education object for each association rule and a second voting proportion of an expert object for each association rule based on the feedback data; and determine the interest degree of each association rule in the set of association rules based on the first voting proportion and the second voting proportion.

[0136] In some embodiments of the disclosure, the updating unit 404 is configured to: filter a set of to-be-transformed rules from the set of association rules based on the interest degree; filter matching results related to the set of to-be-transformed rules from the matching result; transform the set of to-be-transformed rules into information tuples based on the matching results of the set of to-be-transformed rules; and update the information tuples in the knowledge base by using the transformed information tuples.

[0137] In some embodiments of the present disclosure, the updating unit 404 is further configured to: add the transformed information tuple to the knowledge base, and remove the repeated information tuples in the knowledge base; collect the prediction accuracy of each information tuple in the knowledge base; determine the relevance of each information tuple; determine the priority of each information tuple based on the prediction accuracy and the relevance; and sort the information tuples in the knowledge base based on the priority of each information tuple.

[0138] In some embodiments of the present disclosure, the device further comprises a constructing unit (not shown in the figure) configured to: determine the association rule; determine the question in the teaching scene or the course design based on the association rule; determine the recommended information based on the question, and send the question and the recommended information; obtain the relevance of the question and the recommended information; and take the association rule, the question, the recommended information, and the relevance as an information tuple.

[0139] In some embodiments of the present disclosure, the recommendation unit 405 is further configured to: determine the order of each information tuple based on the information tuples in the knowledge base; determine the information tuple to be recommended based on the order; extract the recommended information in the information tuple to be recommended, and send the recommended information to the education object; receive the recommended course sent by the education object, and send the recommended course to the learning object.

[0140] The information recommendation device provided by the embodiments of the present disclosure first determines and sends the association rule set based on the data set when the learning object completes the initial course by the determining unit 401; secondly, the interest degree of the user to the association rule set is obtained by the obtaining unit 402; thirdly, the association rule set is matched with the rules in the knowledge base to determine the matching result of the association rule set by the matching unit 403; then, the information tuple in the knowledge base is updated based on the interest degree, the matching result, and the association rule set by the updating unit 404; finally, the recommended course is determined based on the information tuples in the knowledge base, and the recommended course is sent to the learning object by the recommendation unit 405. Thus, by matching the association rule set with the rules in the knowledge base and updating the information tuples in the knowledge base, the dynamic and personalized course recommendation is realized through the information tuples, the reliability and accuracy of the recommended course are improved, and the experience of the learning object is improved. At the same time, the continuous updating of the knowledge base also ensures the continuous optimization and adaptability of the recommendation system.

[0141] Further reference Figure 5 , as an implementation of the method shown in the above figures, the present disclosure provides an embodiment of an information recommendation system, and the device embodiment corresponds to the method embodiment shown in Figure 1 .

[0142] As shown in Figure 5 , the information recommendation system 500 provided by the present embodiment comprises a client 501 and a server 502.

[0143] The client 501 is configured to determine and send a set of association rules based on a data set when a learning object completes an initial course, acquire an interest degree of the learning object on the set of association rules, match the set of association rules with rules in the knowledge base 5011 to determine a matching result of the set of association rules, update information tuples in the knowledge base 5011 based on the interest degree, the matching result and the set of association rules, and determine a recommended course based on the information tuples in the knowledge base 5011 and send the recommended course to the learning object. The client 501 can be a terminal used by an education object (such as a teacher, a training subject, etc.), which applies the association rule mining algorithm locally and is part of a course development iteration process, and can be used to find problems and feedback, and the object of the problems and feedback can be an object educated by the education object, such as the learning object. The problem can be the recommended course recommended to the learning object. The client can download the knowledge base, receive an update of the knowledge base, and apply the mining algorithm to the knowledge base in an offline state to update the knowledge base.

[0144] The server 502 includes a network application module 5021 and a network service module 5022. The network application module 5021 is configured to provide an interface for managing the knowledge base 5011, so as to update the knowledge base 5011 through the interface. The network service module 5022 is configured to provide the updated knowledge base to the client 501. The interface provided by the network application module 5021 can be used by an expert to manage the knowledge base in a manner of adding, deleting, modifying, searching, voting, etc., so as to achieve the purpose of associating the knowledge base. After receiving the updated knowledge base, the client 501 manages the local knowledge base 5011 in the same manner as the received updated knowledge base. In this embodiment, the network service module 5022 uses PMML (Predictive Model Markup Language) to share the knowledge base updated through the network application module 5021 to the client 501. The use of PMML in the network service module 5022 ensures the standardization and interoperability of model sharing, and avoids compatibility problems.

[0145] In this embodiment, the specific processing of the client 501 and the technical effects brought by the specific processing can be respectively referred to the related description of the step 101, the step 102, the step 103, the step 104 and the step 105 in the corresponding embodiment, which will not be described herein again. Figure 1 The corresponding embodiment corresponds to the step 101, the step 102, the step 103, the step 104 and the step 105, and the related description of the step 101, the step 102, the step 103, the step 104 and the step 105 will not be described herein again.

[0146] The information recommendation system provided by the embodiment discovers a pattern through data mining of the client, combines the knowledge of experts and teachers to evaluate and refine the knowledge base, updates the knowledge base, combines the knowledge of the updated knowledge base, and finally feeds back the recommended course to the learning object in the form of recommendation, forms a closed loop, and realizes continuous improvement of the electronic learning course.

[0147] According to embodiments of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium and a computer program product.

[0148] Figure 6 A schematic block diagram of an example electronic device 600 that can be used to implement embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptops, desktops, tablets, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular telephones, smart phones, wearable devices, and other similar computing devices. The components shown here, their connections and relationships, and their modes of operation, are meant to be examples only, and are not meant to limit the implementations of the present disclosure described and / or claimed in this document.

[0149] As shown in Figure 6 The electronic device 600 includes a computing unit 601 that can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 602 or a computer program loaded from a storage unit 608 into a random access memory (RAM) 603. Various programs and data required for the operation of the electronic device 600 can also be stored in the RAM 603. The computing unit 601, the ROM 602, and the RAM 603 are connected to each other through a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0150] Various components in the electronic device 600 are connected to the I / O interface 605, including an input unit 606, such as a keyboard, a mouse, etc., an output unit 607, such as various types of displays, a speaker, etc., a storage unit 608, such as a magnetic disk, an optical disk, etc., and a communication unit 609, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 609 allows the electronic device 600 to exchange information / data with other devices through a computer network, such as the Internet, and / or various telecommunication networks.

[0151] The computing unit 601 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 601 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various specialized artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, and the like. The computing unit 601 performs various methods and processes described above, such as the information recommendation method. For example, in some embodiments, the information recommendation method can be implemented as a computer software program tangibly embodied in a machine-readable medium, such as the storage unit 608. In some embodiments, part or all of the computer program can be loaded and / or installed onto the device 600 via the ROM 602 and / or the communication unit 609. When the computer program is loaded onto the RAM 603 and executed by the computing unit 601, one or more steps of the information recommendation method described above can be performed. Alternatively, in other embodiments, the computing unit 601 can be configured to perform the information recommendation method by any other suitable means, such as by means of firmware.

[0152] Various implementations of the systems and techniques described above can be realized in digital electronic circuitry, integrated circuitry, specially designed application specific integrated circuits (ASICs), field programmable gate arrays (FPGAs) computer hardware, firmware, software, and / or combinations thereof. These various implementations can include implementation in one or more computer programs that are executable and / or interpretable on a programmable system including at least one programmable processor, which can be special or general purpose, coupled to receive data and instructions from, and to transmit data and instructions to, a storage system, at least one input device, and at least one output device.

[0153] Program code for carrying out methods of the present disclosure can be written in any combination of one or more programming languages. The program code can be provided to a processor or controller of a general purpose computer, special purpose computer, or other programmable information recommendation apparatus to produce a machine, such that the program code, when executed by the processor or controller, implements the methods / operations specified in the flowcharts and / or block diagrams. The program code can execute entirely on a machine, partly on the machine, as a stand-alone software package, partly on the machine and partly on a remote machine or entirely on the remote machine or server.

[0154] In the context of this disclosure, a machine-readable medium can be a tangible medium that contains or stores a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can include but is not limited to an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of the machine-readable storage medium would include a lined- up electrical connection, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0155] To provide for interaction with a user, the systems and techniques described here can be implemented on a computer having a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the computer. Other kinds of devices can be used to provide for interaction with a user as well; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form, including acoustic, speech, or tactile input.

[0156] The systems and techniques described here can be implemented in a computing system that includes a back end component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front end component (e.g., a user computer having a graphical user interface or a Web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such back end, middleware, or front end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network (LAN), a wide area network (WAN), and the Internet.

[0157] It should be understood that various forms of flow shown above can be re-ordered, added to, or deleted from without departing from the spirit of the present disclosure. For example, the steps recited in the present disclosure can be performed in parallel, in series, or in a different order, without departing from the desired results of the technology disclosed in the present disclosure, which are not limited herein.

[0158] The foregoing description of specific exemplary embodiments of the disclosure has been presented for the purposes of illustration and explanation. It is not intended to be exhaustive or to limit the disclosure to the precise forms disclosed, and obviously many modifications and variations are possible in light of the above teaching. It is intended that the embodiments be chosen and / or described such that the disclosure is practical and sufficient for one skilled in the art to practice and utilize the disclosure, as well as to customize the disclosure for various uses and conditions. The scope of the disclosure is intended to be limited only by the claims and their equivalents.

Claims

1. An information recommendation method, the method comprising: determining and sending a set of association rules based on a data set when a learning object completes an initial course; obtaining a degree of interest of a user in the set of association rules; matching the set of association rules with rules in a knowledge base to determine a matching result of the set of association rules; updating information tuples in the knowledge base based on the degree of interest, the matching result and the set of association rules; the updating of the information tuples in the knowledge base based on the degree of interest, the matching result and the set of association rules comprises: screening a set of to-be-transformed rules from the set of association rules based on the degree of interest; screening a matching result related to the set of to-be-transformed rules from the matching result; transforming the set of to-be-transformed rules into information tuples based on the matching result of the set of to-be-transformed rules; and updating the information tuples in the knowledge base by using the transformed information tuples; determining a recommended course based on the information tuples in the knowledge base and sending the recommended course to the learning object; the method further comprises: obtaining a data set when the learning object completes the recommended course; determining and sending a new set of association rules based on the data set when the learning object completes the recommended course; generating a new recommended course based on the new set of association rules, a degree of interest of a user in the new set of association rules and a matching result of the new set of association rules in the knowledge base, and sending the new recommended course to the learning object.

2. The method of claim 1, wherein, the determining and sending of the set of association rules based on the data set when the learning object completes the initial course comprises: preprocessing the data set when the learning object completes the initial course to obtain preprocessed data; performing rule analysis on the preprocessed data and a search area selected by an educational object by using an association analysis algorithm to obtain the set of association rules; sending the set of association rules to the educational object.

3. The method of claim 1, wherein, the obtaining of the degree of interest of the user in the set of association rules comprises: obtaining feedback data of the user in each association rule in the set of association rules; determining a first voting proportion of the educational object in each association rule and a second voting proportion of an expert object in each association rule based on the feedback data; determining a degree of interest of each association rule in the set of association rules based on the first voting proportion and the second voting proportion.

4. The method of claim 3, wherein, the determining of the degree of interest of each association rule in the set of association rules based on the first voting proportion and the second voting proportion comprises: determining a first weighting coefficient of the educational object and a second weighting coefficient of the expert object; obtaining a weighted measurement value based on the first voting proportion, the second voting proportion, the first weighting coefficient and the second weighting coefficient; performing accuracy prediction on a voting result of the educational object in each association rule to obtain an accuracy result of each association rule; obtaining the degree of interest of each association rule based on the weighted measurement value and the accuracy result.

5. The method of claim 1, wherein, the updating of the information tuples in the knowledge base by using the transformed information tuples comprises: adding the transformed information tuples to the knowledge base and removing duplicate information tuples in the knowledge base; collecting a prediction accuracy of each information tuple in the knowledge base; determining a relevance of each information tuple; determining a priority of each information tuple based on the prediction accuracy and the relevance; ordering the information tuples in the knowledge base based on the priority of each information tuple.

6. The method of any of claims 1-4, wherein, The information tuples in the knowledge base are constructed by the following steps: obtaining a cold start rule of an expert object; determining an expert tuple based on the cold start rule; obtaining an education tuple sent by an education object; sending the education tuple to the expert object; in response to receiving a verification pass result of the expert object on the education tuple, taking both the expert tuple and the education tuple as information tuples in the knowledge base.

7. The method of any one of claims 1-4, wherein, The information tuples in the knowledge base are also constructed by the following steps: determining an association rule; determining a problem in a teaching scenario or course design based on the association rule; determining recommended information based on the problem and sending the problem and the recommended information; obtaining a relevance of the problem and the recommended information; taking the association rule, the problem, the recommended information, and the relevance as an information tuple.

8. The method of claim 1, wherein, The determining of the recommended course based on the information tuples in the knowledge base and the sending of the recommended course to the learning object include: determining an ordering of each information tuple based on the information tuples in the knowledge base; determining an information tuple to be recommended based on the ordering; extracting recommended information in the information tuple to be recommended and sending the recommended information to an education object; receiving a recommended course sent by the education object and sending the recommended course to the learning object.

9. An information recommendation apparatus, the apparatus comprising: a determination unit configured to determine and send a set of association rules based on a data set when a learning object completes an initial course; an obtaining unit configured to obtain an interest degree of a user on the set of association rules; a matching unit configured to match the set of association rules with rules in a knowledge base to determine a matching result of the set of association rules; an updating unit configured to update information tuples in the knowledge base based on the interest degree, the matching result, and the set of association rules; the updating unit is configured to: filter a set of rules to be converted from the set of association rules based on the interest degree; filter a matching result related to the set of rules to be converted from the matching result; convert the set of rules to be converted into information tuples based on the matching result of the set of rules to be converted; and update the information tuples in the knowledge base by using the converted information tuples; a recommendation unit configured to determine a recommended course based on the information tuples in the knowledge base and send the recommended course to the learning object. The device is further configured to: acquire a data set of the learning object completing the recommended course; determine and send a new set of association rules based on the data set of the learning object completing the recommended course; generate a new recommended course based on the new set of association rules, the interest degree of the user in the new set of association rules, and the matching result of the new set of association rules in the knowledge base, and send the new recommended course to the learning object.

10. An information recommendation system, the system comprising: A client and a server; The client is configured to determine and send a set of association rules based on a data set of the learning object completing an initial course; Acquire the interest degree of the user in the set of association rules; Match the set of association rules with rules in a knowledge base to determine a matching result of the set of association rules; Update information tuples in the knowledge base based on the interest degree, the matching result, and the set of association rules; Determine a recommended course based on the information tuples in the knowledge base and send the recommended course to the learning object; The updating of the information tuples in the knowledge base based on the interest degree, the matching result, and the set of association rules includes: filtering a set of to-be-transformed rules from the set of association rules based on the interest degree; filtering matching results related to the set of to-be-transformed rules from the matching result; transforming the set of to-be-transformed rules into information tuples based on the matching result of the set of to-be-transformed rules; and updating the information tuples in the knowledge base by using the transformed information tuples; the client is further configured to: acquire a data set of the learning object completing the recommended course; determine and send a new set of association rules based on the data set of the learning object completing the recommended course; generate a new recommended course based on the new set of association rules, the interest degree of the user in the new set of association rules, and the matching result of the new set of association rules in the knowledge base, and send the new recommended course to the learning object; The server includes: a network application module and a network service module, the network application module is configured to provide an interface for managing the knowledge base, so as to update the knowledge base through the interface; and the network service module is configured to provide the updated knowledge base to the client.

11. An electronic device, comprising: Comprise: At least one processor; And a memory connected with the at least one processor in communication; Wherein, the memory stores instructions executable by the at least one processor, the instructions are executed by the at least one processor to enable the at least one processor to execute the method of any one of claims 1-8.

12. A non-transitory computer-readable storage medium having stored thereon computer instructions, wherein, The computer instructions are used to enable the computer to execute the method of any one of claims 1-8. The computer instructions are used to enable the computer to execute the method of any one of claims 1-8.

Images