A big data-based teaching resource intelligent recommendation method

Abstract

The present application relates to the technical field of teaching resource recommendation, and particularly relates to a teaching resource intelligent recommendation method based on big data. The method comprises the following steps: obtaining a user set of a current user, a label of each teaching resource, and a browsing frequency of each user in the user set to each label; obtaining an occurrence rate of each user and each label and an average occurrence rate of each label to obtain an average attention degree of the label; obtaining a to-be-analyzed label set of the current user according to the average attention degree of each label; obtaining a single degree of each to-be-analyzed label in the to-be-analyzed label set; then obtaining a single trend performance of each to-be-analyzed label; then obtaining an interesting degree of each to-be-analyzed label; finally obtaining an importance degree of the to-be-analyzed label; obtaining a priority of each to-be-analyzed label; then obtaining a recommendation degree of each teaching resource; and recommending each teaching resource to the current user in a descending order of the recommendation degree. The present application can effectively recommend the required teaching resource to the user.

Description

Technical Field

[0001] This invention relates to the field of teaching resource recommendation technology, specifically to an intelligent teaching resource recommendation method based on big data. Background Technology

[0002] With the development of big data technology, teaching resources are becoming increasingly abundant and diverse. To improve the efficiency of resource utilization and enhance students' personalized learning experience, intelligent recommendation of teaching resources has become an important part of online education platforms and other learning systems. Improving the alignment of recommendation results with users' learning needs and interests has become a key development direction for the informatization of teaching.

[0003] Existing intelligent recommendation methods for teaching resources mainly recommend resources with the same tags by statistically analyzing users' historical browsing and learning of teaching resources (such as category tags, knowledge point tags, etc.). However, traditional methods rely solely on statistical results for recommendations, lacking analysis of the distribution of each tag in historical records and the dynamic changes in users' learning needs. This can easily lead to a gradual homogenization of recommended content over time, making it difficult to comprehensively cover users' learning needs. Summary of the Invention

[0004] To address the aforementioned technical problems, the present invention aims to provide an intelligent recommendation method for teaching resources based on big data, the specific technical solution of which is as follows: One embodiment of the present invention provides an intelligent recommendation method for teaching resources based on big data, the method comprising: The current user and other users with the same user information are grouped into a user set for the current user; the tags for each teaching resource and the number of times each user in the user set viewed each tag within the historical time period are obtained; The occurrence rate of a user for a given tag is obtained by comparing the number of times a user views a tag with the total number of times that user views all tags in the user set. The average occurrence rate of each tag is then calculated. The average attention level of a tag is obtained by combining the average occurrence rate of a tag with the number of users in the user set who have viewed that tag. The set of tags to be analyzed for the current user is then obtained based on the average attention level of each tag. The degree of specialization of each tag to be analyzed is obtained by calculating the occurrence rate of each tag for the current user in the tag set to be analyzed; the specialization trend of a tag to be analyzed is obtained by calculating the degree of specialization of a tag to be analyzed, the total number of times the current user views the tag, and the average total number of times all users in the user set view the tag; the degree of interest in each tag to be analyzed is obtained by calculating the occurrence rate of each tag for the current user each day; the importance of a tag to be analyzed is obtained by adding the normalized value of the average attention of a tag to be analyzed to the degree of interest and normalizing the result. The priority of a tag to be analyzed is obtained based on its importance and trend of homogenization; the priority of each tag to be analyzed corresponding to a teaching resource is added together to obtain the recommendation level of the teaching resource; and the teaching resources are recommended to the current user in descending order of recommendation level.

[0005] Preferably, the occurrence rate of a user for a given tag is obtained based on the number of times a user views a tag within the user set and the total number of times that user views all tags, and the average occurrence rate of all tags is then calculated, including: The occurrence rate of a user for a given tag is obtained by comparing the number of times a user views a tag with the total number of times the user views all tags. The average occurrence rate of a tag is calculated as the average occurrence rate of the tag for all users in the user set.

[0006] Preferably, the average attention level of a tag is obtained by using the average occurrence rate of a tag and the number of users in the user set who have viewed the tag, including: The coverage of a tag is obtained by comparing the number of users in the user set who have viewed the tag with the total number of users in the user set; the average attention of a tag is obtained by multiplying the coverage of the tag by the average attention of the tag.

[0007] Preferably, the set of tags to be analyzed for the current user is obtained based on the average attention level of each tag, including: The k-means clustering method is used to cluster the average attention of each tag viewed by all users in the user set to obtain different clusters; the average value of the average attention of each tag in each cluster is obtained, and the tags corresponding to the cluster with the largest average value and the tags viewed by the current user form the tag set to be analyzed for the current user.

[0008] Preferably, the degree of uniformity of each tag to be analyzed is obtained based on the occurrence rate of each tag to be analyzed for the current user in the tag set to be analyzed, including: The average of the differences between the occurrence rate of a tag to be analyzed for the current user and the occurrence rates of all other tags to be analyzed for the current user in the tag set to be analyzed is obtained and normalized to obtain the degree of uniformity of the tag to be analyzed.

[0009] Preferably, the uniformity trend of a tag to be analyzed is obtained based on the degree of uniformity of the tag, the total number of times the current user views the tag, and the average of the total number of times all users in the user set view the tag, including: The credibility of the current user's tag repetition performance is obtained by comparing the total number of times the current user browses tags with the mean of the total number of times all users in the user set browse tags and normalizing the result. The credibility of the current user's tag repetition performance is then multiplied by the degree of uniformity of a tag to be analyzed to obtain the uniformity trend of that tag.

[0010] Preferably, the degree of interest in each tag to be analyzed is obtained based on the occurrence rate of each tag to be analyzed by the current user on each day, including: The slope value of the fitted line for the tag to be analyzed is obtained by linearly fitting the occurrence rate of the current user for each day, and the slope value is normalized to obtain the interest level of the tag to be analyzed.

[0011] Preferably, the priority of the tag to be analyzed is obtained based on its importance and homogenization trend, including: The difference between a first preset value and the simplification value of a tag to be analyzed is obtained and multiplied by the importance of the tag to be analyzed to obtain the priority of the tag to be analyzed.

[0012] The embodiments of the present invention have at least the following beneficial effects: This application forms a user set for the current user and other users with the same user information as the current user; obtains the tags of each teaching resource and the number of times each user in the user set viewed each tag in the historical period; then analyzes the number of times the tags of users in the user set of the same type as the current user to obtain the tag set to be analyzed for the current user, thereby accurately identifying the tags of teaching resources that are more suitable for the current user; Furthermore, the degree of homogeneity of each tag to be analyzed is obtained based on the occurrence rate of each tag for the current user in the tag set to be analyzed; the homogeneity trend of a tag to be analyzed is obtained based on the degree of homogeneity of a tag to be analyzed, the total number of times the current user browses the tag, and the average of the total number of times all users in the user set browse the tag. This analysis helps to determine whether the current user has a trend of browsing resources homogeneity, thereby reducing the priority of tags with excessively high occurrence frequency in recommendations. This can effectively reduce the trend of high repetition and homogeneity in recommendation results, which is conducive to users' more comprehensive and diverse learning. Next, the interest level of each tag to be analyzed is obtained based on the occurrence rate of each tag for the current user on each day. The normalized value of the average attention of a tag to be analyzed is added to the interest level and normalized to obtain the importance of the tag to be analyzed. Here, based on the current user's interest bias and the attention level of similar users with the same user information to each tag, the importance of each tag is analyzed, and the priority of tags with higher importance is increased. This can effectively avoid some historically repetitive but important content being ignored, making the recommendation results more in line with the current user's learning needs. Attached Figure Description

[0013] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0014] Figure 1 This application provides a flowchart of a method for intelligent recommendation of teaching resources based on big data. Detailed Implementation

[0015] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a big data-based intelligent recommendation method for teaching resources proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0016] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0017] The following description, in conjunction with the accompanying drawings, details a specific scheme for an intelligent recommendation method for teaching resources based on big data, provided by this invention.

[0018] Example: The main application scenario of this invention is as follows: When users browse learning resources on online education platforms and other learning systems, they tend to browse more of the teaching resources that interest them. Therefore, traditional teaching resource recommendation methods often rely on statistically analyzing the number of times each teaching resource's tags have been viewed historically. However, this method leads to repeated recommendations of teaching resources corresponding to certain tags, resulting in increasingly homogenous recommendations that fail to meet users' actual learning needs. Therefore, it is necessary to perform relevant analysis on users' historical browsing data to recommend the learning resources they require.

[0019] Please see Figure 1 The diagram illustrates a flowchart of an intelligent recommendation method for teaching resources based on big data, provided by an embodiment of the present invention. The method includes the following steps: Step S1: Combine the current user and other users with the same user information into a user set for the current user; obtain the tags for each teaching resource and the number of times each user in the user set viewed each tag within the historical time period.

[0020] The user's learning background is a major constraint on the direction of teaching resource recommendations. Therefore, the first step is to obtain the user's information, including their grade level, major, curriculum, and learning goals. Users with the same information may choose different teaching resources due to different learning preferences. Therefore, the number of times a user has viewed the tags of teaching resources in the past and the corresponding tags of the teaching resources are obtained. This application obtains tags for each learning resource, including subject category tags, knowledge point tags, and resource format tags. Specific tag types can be labeled according to actual needs, and a teaching resource may have multiple tags. This application collects the number of times each user viewed each tag within a historical period, specifically the most recent three months. Furthermore, users may change their information (grade level, major, curriculum, learning goals) during this period. To avoid the influence of data before the user type change, if user information has changed, only the historical data after the information change is obtained; that is, only the number of times the user viewed each tag after the information change is obtained.

[0021] Furthermore, in order to perform more accurate analysis by combining users with the same user information, the current user and other users with the same user information are combined into a user set for the current user.

[0022] Step S2: Obtain the occurrence rate of a user for a given tag based on the number of times a user views a tag in the user set and the total number of times the user views all tags, and obtain the average occurrence rate of each tag; use the average occurrence rate of a tag and the number of users in the user set who have viewed the tag to obtain the average attention level of the tag; obtain the set of tags to be analyzed for the current user based on the average attention level of each tag.

[0023] Because when a user first starts using the current educational platform or enters a new learning stage, due to insufficient historical information available, it may be difficult to accurately obtain the tags of teaching resources that match the current user. However, users in the user set (with the same user information) have similar main learning scopes. Therefore, by comparing the main tags of teaching resources browsed by other users in the current user's user information history, a set of tags to be analyzed can be obtained.

[0024] The more times a tag appears in a user's browsing history, the higher the user's attention to that tag, meaning it may be more important to that user. Therefore, the occurrence rate of a tag for a user can be obtained by comparing the number of times a user browses a tag with the total number of times the user browses all tags in the user set.

[0025] Specifically, the frequency of a user viewing a particular tag is calculated by comparing the number of times a user views a tag with the total number of times that user views all tags.

[0026] The specific model for calculating the occurrence rate of a user with a single tag is as follows: , in, This represents the occurrence rate of the i-th tag in the historical data of the u-th user, which is also the occurrence rate of the i-th tag for the u-th user. This represents the number of times the u-th user views the i-th tag, which is also the total number of times the u-th user views the i-th tag. This represents the total number of times the u-th user has viewed each tag throughout history. Furthermore, for a given tag, the average occurrence rate of that tag is calculated as the mean occurrence rate for that tag across all users in the user set.

[0027] However, due to different user preferences, some tags may only appear frequently in the historical data of some users. These tags cannot reflect the general learning direction of such users. Therefore, in order to avoid the high average occurrence rate of such tags due to the influence of some users' own learning preferences, it is also necessary to calculate the coverage of each tag in the user browsing data (i.e., the proportion of users who have browsed the tag among all similar users).

[0028] The average attention level of a tag can be obtained by using its average occurrence rate and the number of users in the user set who have viewed the tag.

[0029] Specifically, the coverage of a tag is obtained by comparing the number of users in the user set who have viewed the tag with the total number of users in the user set; the average attention of a tag is obtained by multiplying the coverage of the tag by the average attention of the tag.

[0030] The specific model for calculating the average attention of a tag is as follows: , in, Let be the average attention level of the i-th label, representing the average level of attention that this type of user (users in the user set) pays to the i-th label; Let be the average occurrence rate of the i-th label, representing the average occurrence rate of the i-th label in the historical data of users in the user set. This represents the coverage of the i-th tag in the browsing data of this type of user.

[0031] To obtain the set of tags that all users in the user set primarily focus on, the k-means clustering method is used to calculate the average attention level of each tag viewed by all users in the user set. Clustering is performed (where the k-value is determined using the elbow method) to obtain different clusters, and the average attention value of each tag in each cluster is obtained. ,in The largest cluster corresponds to the tags that this type of user (user set) primarily focuses on. Therefore, obtaining... The tags in the largest cluster, along with the tags that the current user has viewed in the past, form the tag set to be analyzed for the current user.

[0032] Step S3: Obtain the degree of uniformity of each tag to be analyzed based on the occurrence rate of each tag to be analyzed for the current user in the tag set to be analyzed; obtain the uniformity trend of the tag to be analyzed based on the degree of uniformity of the tag to be analyzed, the total number of times the current user browses the tag, and the average of the total number of times all users in the user set browse the tag; obtain the degree of interest of each tag to be analyzed based on the occurrence rate of each tag to be analyzed for the current user each day; add the normalized value of the average attention of the tag to be analyzed to the degree of interest and normalize it to obtain the importance of the tag to be analyzed.

[0033] Traditional recommendation methods often repeatedly suggest frequently viewed educational resource tags, making it harder for users to find resources they might actually need. Therefore, this study analyzes the repetitive nature of each tag in the current user's historical browsing data to determine the likelihood of a user exhibiting a tendency to focus on a single type of educational resource. This serves as one of the criteria for prioritizing each tag, appropriately reducing the weight of excessively repetitive tags in the recommendation results.

[0034] However, a user's browsing of educational resource tags is related to their current learning objectives. This means that some tags may have high repetition in historical data, but the content corresponding to those tags (such as knowledge points) may be highly important to the current user. Therefore, adjusting based solely on repetition may miss some important educational resources, resulting in recommendations that do not meet user needs. Thus, in addition to the repetition index, it is also necessary to calculate the priority of each tag based on its importance to the current user.

[0035] A single teaching resource may correspond to multiple tags. This means that if multiple tags have relatively high occurrence rates, the teaching resource still exhibits high diversity. Therefore, it is necessary to compare the differences in the occurrence rates of different tags within the set to be analyzed, and calculate the occurrence rate of each tag.

[0036] The degree of uniformity of each tag to be analyzed is obtained by calculating the occurrence rate of each tag for the current user in the tag set to be analyzed. Specifically, the average difference between the occurrence rate of one tag for the current user and the occurrence rates of all other tags for the current user in the tag set to be analyzed is calculated and normalized to obtain the degree of uniformity of that tag.

[0037] The specific calculation model for the degree of uniqueness of a tag to be analyzed is as follows: , in, The degree of uniformity of the i-th tag to be analyzed represents the degree of uniformity of the i-th tag to be analyzed in the current user's browsing history; norm represents the normalization function; This represents the total number of tags to be analyzed in the set of tags to be analyzed. This represents the occurrence rate of the i-th tag to be analyzed for the current user, which is also the occurrence rate in the current user's historical data. This represents the occurrence rate of the j-th tag to be analyzed for the current user. It should be noted that there may be tags in the set of tags to be analyzed that are not in the current user's browsing history. In this case, the occurrence rate of the tag to be analyzed is recorded as 0. The higher the occurrence rate of the i-th tag to be analyzed compared to other tags to be analyzed, the more likely the tag to be analyzed is to reflect a trend of homogenization in the current user's historical browsing resources.

[0038] Since the historical data corresponding to the current user may be small due to recent changes in user information, the repetitive performance of each tag to be analyzed cannot effectively reflect whether each tag to be analyzed has a single performance in the historical data. Therefore, the credibility of the repetitive performance of the tags to be analyzed is calculated based on the total number of historical views of the current user, and then combined with the degree of singleness of the tags to be analyzed to obtain the singleness trend performance of the tags to be analyzed.

[0039] The uniformity trend of a tag to be analyzed is obtained by considering the degree of uniformity of the tag, the total number of times the current user views the tag, and the average total number of times all users in the user set view the tag.

[0040] Specifically, the credibility of the current user's tag repetition performance is obtained by comparing the total number of times the current user browses tags with the average total number of times all users in the user set browse tags and normalizing the result; the credibility of the current user's tag repetition performance is then multiplied by the degree of uniformity of a tag to be analyzed to obtain the uniformity trend of that tag.

[0041] The specific calculation model for the uniformity trend of a label to be analyzed is as follows: , in, This represents the homogenization trend of the i-th tag to be analyzed, reflecting the homogenization trend of the i-th tag to be analyzed in the current user's historical data. This indicates the degree of uniformity of the i-th label to be analyzed; This indicates the total number of times the current user has viewed the tags. This represents the average total number of times each user in the user set has viewed a tag. The more tags a current user has viewed, the higher the reliability. This represents the credibility of the current user's tag repetition; norm is the normalization function.

[0042] Although similar users (users in a user set) may pay attention to similar tags, different users may have different personal interests and preferences. Therefore, the tags that the current user prefers should be more important.

[0043] The degree of interest in each tag to be analyzed is obtained based on the occurrence rate of each tag for the current user on each day. Specifically, a linear fit is performed on the occurrence rate of a tag to be analyzed for the current user on each day to obtain the slope value of the fitted line for that tag, and the slope value is normalized to obtain the degree of interest in that tag.

[0044] The occurrence rate of a single tag to be analyzed for a current user per day is the ratio of the number of times the current user viewed that tag that day to the total number of times the current user viewed all tags that day. A larger slope value for a single tag indicates a greater increasing interest from the current user in that tag. Furthermore, if the current user's historical browsing data is less than two days, trend analysis is not possible; therefore, the interest level of other users in the user set for each tag to be analyzed is calculated using the method described above. (The degree of interest of the u-th user corresponding to the i-th tag), with The average value is used as the interest level corresponding to the i-th tag to be analyzed for the current user. .

[0045] Next, by combining the average attention given to each tag to be analyzed by users in the user set and the current user's interest in each tag to be analyzed, the importance of each tag to be analyzed to the current user is calculated: , in, This indicates the importance of the i-th tag to be analyzed for the current user. This represents the normalized value of the average attention received by the i-th tag to be analyzed. This represents the level of interest in the i-th label. `norm` represents the normalization function. This allows us to obtain the importance of each label to be analyzed.

[0046] Step S4: Obtain the priority of the tag to be analyzed based on its importance and homogenization trend; sum the priorities of each tag to be analyzed corresponding to a teaching resource to obtain the recommendation level of the teaching resource; recommend each teaching resource to the current user in descending order of recommendation level.

[0047] The above describes the individual performance trend and importance of each tag to be analyzed. Combining these two factors, the priority of each tag can be determined. Therefore, the priority of a tag to be analyzed is obtained based on its importance and individual performance trend.

[0048] Specifically, the difference between a first preset value and a single performance value of a tag to be analyzed is obtained and multiplied by the importance of the tag to be analyzed to obtain the priority of the tag to be analyzed.

[0049] The specific calculation model for the priority of a tag to be analyzed is as follows: , in, This indicates the priority of the i-th tag to be analyzed, reflecting the priority of the i-th tag to the current user. This indicates the importance of the i-th label to be analyzed. This represents the uniformity trend of the i-th tag to be analyzed. The first preset value is 1.

[0050] Furthermore, since each teaching resource has multiple tags, the recommendation level of each teaching resource is calculated based on the sum of the priority of the tags corresponding to each teaching resource.

[0051] Specifically, the recommendation level of a teaching resource is obtained by summing the priorities of each tag to be analyzed; the calculation model for the recommendation level of a teaching resource is as follows: , This represents the recommendation level of the e-th teaching resource, reflecting the degree to which the e-th teaching resource is recommended to the current user. This represents the total number of tags to be analyzed corresponding to the e-th teaching resource. This represents the priority of the i-th tag to be analyzed corresponding to the e-th teaching resource for the current user. During the calculation, some tags corresponding to a teaching resource may not be in the set of tags to be analyzed; these tags are not included in the recommendation calculation. Alternatively, if they are to be included in the calculation, their priority can be set to 0.

[0052] Finally, the teaching resources were ranked according to their recommendation level. The resources are sorted from highest to lowest quality and recommended to the current user in that order. For example, if the online education platform's recommendation interface can recommend 10 teaching resources at a time, then those resources will be displayed on the recommendation interface. The top 10 teaching resources.

[0053] In summary, this application, based on statistical analysis of users' historical browsing history of teaching resource tags, analyzes the repetition rate of each tag according to its distribution in historical data, and analyzes the importance of each tag to the current user based on the trend of its occurrence rate. This allows for the adjustment of the priority of each tag, resulting in a recommendation order for teaching resources based on the priority of the tags corresponding to each resource. By dynamically adjusting the weight of teaching resource tags, the problem of homogeneous recommended content can be effectively reduced, leading to more comprehensive recommendation results.

[0054] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0055] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0056] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for intelligent recommendation of teaching resources based on big data, characterized in that, The method includes: The current user and other users with the same user information are grouped into a user set for the current user; the tags for each teaching resource and the number of times each user in the user set viewed each tag within the historical time period are obtained; The occurrence rate of a user for a given tag is obtained by comparing the number of times a user views a tag with the total number of times that user views all tags in the user set. The average occurrence rate of each tag is then calculated. The average attention level of a tag is obtained by combining the average occurrence rate of a tag with the number of users in the user set who have viewed that tag. The set of tags to be analyzed for the current user is then obtained based on the average attention level of each tag. The degree of specialization of each tag to be analyzed is obtained by calculating the occurrence rate of each tag for the current user in the tag set to be analyzed; the specialization trend of a tag to be analyzed is obtained by calculating the degree of specialization of a tag to be analyzed, the total number of times the current user views the tag, and the average total number of times all users in the user set view the tag; the degree of interest in each tag to be analyzed is obtained by calculating the occurrence rate of each tag for the current user each day; the importance of a tag to be analyzed is obtained by adding the normalized value of the average attention of a tag to be analyzed to the degree of interest and normalizing the result. The priority of a tag to be analyzed is obtained based on its importance and trend of homogenization; the priority of each tag to be analyzed corresponding to a teaching resource is added together to obtain the recommendation level of the teaching resource; and the teaching resources are recommended to the current user in descending order of recommendation level.

2. The intelligent recommendation method for teaching resources based on big data according to claim 1, characterized in that, The step of obtaining the occurrence rate of a user for a specific tag based on the number of times a user views a tag in the user set and the total number of times the user views all tags, and then obtaining the average occurrence rate of all tags, includes: The occurrence rate of a user for a given tag is obtained by comparing the number of times a user views a tag with the total number of times the user views all tags. The average occurrence rate of a tag is calculated as the average occurrence rate of the tag for all users in the user set.

3. The intelligent recommendation method for teaching resources based on big data according to claim 1, characterized in that, The method of obtaining the average attention level of a tag by using its average occurrence rate and the number of users in the user set who have viewed the tag includes: The coverage of a tag is obtained by comparing the number of users in the user set who have viewed the tag with the total number of users in the user set; the average attention of a tag is obtained by multiplying the coverage of the tag by the average attention of the tag.

4. The intelligent recommendation method for teaching resources based on big data according to claim 1, characterized in that, The step of obtaining the set of tags to be analyzed for the current user based on the average attention of each tag includes: The k-means clustering method is used to cluster the average attention of each tag viewed by all users in the user set to obtain different clusters; the average value of the average attention of each tag in each cluster is obtained, and the tags corresponding to the cluster with the largest average value and the tags viewed by the current user form the tag set to be analyzed for the current user.

5. The intelligent recommendation method for teaching resources based on big data according to claim 1, characterized in that, The step of obtaining the degree of uniformity of each tag to be analyzed based on the occurrence rate of each tag for the current user in the tag set to be analyzed includes: The average of the differences between the occurrence rate of a tag to be analyzed for the current user and the occurrence rates of all other tags to be analyzed for the current user in the tag set to be analyzed is obtained and normalized to obtain the degree of uniformity of the tag to be analyzed.

6. The intelligent recommendation method for teaching resources based on big data according to claim 1, characterized in that, The process of obtaining the uniformity trend of a tag to be analyzed based on its degree of uniformity, the total number of times the current user browses the tag, and the average of the total number of times all users in the user set browse the tag includes: The credibility of the current user's tag repetition performance is obtained by comparing the total number of times the current user browses tags with the mean of the total number of times all users in the user set browse tags and normalizing the result. The credibility of the current user's tag repetition performance is then multiplied by the degree of uniformity of a tag to be analyzed to obtain the uniformity trend of that tag.

7. The intelligent recommendation method for teaching resources based on big data according to claim 1, characterized in that, The step of obtaining the interest level of each tag to be analyzed based on the occurrence rate of each tag to be analyzed for each user on each day includes: The slope value of the fitted line for the tag to be analyzed is obtained by linearly fitting the occurrence rate of the current user for each day, and the slope value is normalized to obtain the interest level of the tag to be analyzed.

8. The intelligent recommendation method for teaching resources based on big data according to claim 1, characterized in that, The process of prioritizing a tag based on its importance and homogenization trend includes: The difference between a first preset value and the simplification value of a tag to be analyzed is obtained and multiplied by the importance of the tag to be analyzed to obtain the priority of the tag to be analyzed.

Images