A personalized education recommendation method and system based on youth corpus

Abstract

This invention relates to the field of personalized education technology and discloses a personalized education recommendation method and system based on a corpus of adolescent language, including a data acquisition module and a corpus analysis module. This personalized education recommendation method and system based on a corpus of adolescent language comprehensively collects popular keywords in online language at the current time, individual differences among adolescents, and all learning materials suitable for the current cognitive stage of adolescents through the data acquisition module, and classifies them into datasets. The corpus analysis module analyzes and generates preference data groups, obtaining category tags for each adolescent's browsing preferences, resulting in better personalized education recommendations. The corpus analysis module analyzes and generates recommendation corpus and recommends personalized learning materials to each adolescent's account. Furthermore, it monitors changes in browsing data within an observation period to determine the adolescent's level of interest, adjusts the personalized recommendation strategy accordingly, and flexibly monitors preference changes for more timely optimization and updates.

Description

Technical Field

[0001] The present invention relates to the field of personalized education technology, and in particular to a personalized education recommendation method and system based on a youth corpus. Background Art

[0002] In the field of linguistics and psychology, studying the language usage patterns of people of different ages has important scientific value. The construction of a youth corpus involves a lot of data collection, processing and analysis. Data collection usually uses a variety of methods, including but not limited to diary method, interview method, observation method and modern technology means such as social media data capture.

[0003] At present, the traditional personalized education recommendation method based on youth corpus ignores the differences between individual teenagers and adopts a unified recommendation strategy, which has poor educational learning effects. In addition, in actual use, there is a lack of effective feedback mechanism to collect feedback from teenagers on the recommended content, and it is impossible to optimize and improve the recommendation strategy in a timely manner. Summary of the invention

[0004] 1. Technical issues to be resolved

[0005] In view of the shortcomings of the existing technology, the present invention provides a personalized education recommendation method and system based on a youth corpus, which has the advantages of better personalized education recommendation learning effect, flexible monitoring of preference changes and more timely optimization and updating. It solves the problem that the traditional personalized education recommendation method based on a youth corpus cannot formulate personalized recommendation strategies and lacks a feedback mechanism and optimization and updating are not timely.

[0006] (II) Technical solution

[0007] To achieve the above object, the present invention provides the following technical solution: a personalized education recommendation method based on a youth corpus, comprising the following steps:

[0008] Step 1: The data collection module connects to the big data platform through the network to obtain the language data that appears frequently in the language expression process of teenagers in real time, and uses natural language processing (NLP) technology to pre-process the collected language data, and then extracts keywords from the pre-processed data to form a popular data set;

[0009] Step 2: The data collection module obtains usage records of youth accounts through a network connection to the education platform, and forms an account data set with usage records of all youth accounts;

[0010] Step 3: The data collection module obtains all the material data for recommending young people to study through the Internet connection education platform, and forms all the material data for recommending study into a material data set;

[0011] Step 4: The corpus analysis module generates a preference data set based on the account data set and the material data set. , get the browsing preferences of each youth account;

[0012] Step 5: Corpus analysis module based on preference data group , popular data sets and material data sets, analyze and generate recommended corpora , get the recommended corpus for personalized learning;

[0013] Step 6: Corpus analysis module based on recommended corpus Recommend personalized learning materials to each youth account, and set a fixed observation period for the corpus analysis module , and monitor the observation cycle Based on the changes in the internal account data set, we can determine the degree of interest of teenagers in personalized learning materials and then adjust the personalized recommendation strategy accordingly.

[0014] Preferably, in step one, the preprocessing process includes word segmentation, part-of-speech tagging and syntactic analysis. The word segmentation divides the language data that appears frequently into independent words. The part-of-speech tagging marks the grammatical attributes of the divided words. The syntactic analysis analyzes the structural relationship and overall semantics between each word based on the grammatical attribute markings, and filters out keywords based on the overall semantics of the language data.

[0015] Preferably, in step 1, the expression of the popular data set is , to Corresponding to each high-frequency keyword in the language expression of teenagers, Indicates the time point when the keyword appears frequently, from 1 to The keywords that frequently appear in teenagers’ language expression are indivual.

[0016] Preferably, in step 2, the expression of the account data set is , to The usage records of each youth account are recorded in turn, including the learning materials browsed in the past and the number of times and duration of browsing of each learning material. Indicates the number of times the learning material is viewed in the usage record, from 1 to Indicates the number of youth accounts indivual.

[0017] Preferably, in step 3, the expression of the material data set is , to Corresponding to each material data used to recommend learning to young people, the material data includes learning content, learning form and material evaluation, Indicates the type of learning material, ranging from 1 to Indicates the number of learning materials share.

[0018] Preferably, in step 4, the preference data set The calculation formula is as follows:

[0019] ,

[0020] In the formula, represents the preference data set, Represents all the material data in the comprehensive material dataset used to recommend learning to teenagers, and is marked as , Indicates the account data set The usage records of youth accounts, Indicates that the learning materials browsed in the usage record are substituted into the material dataset for query, and the first Category tags that teenagers browse by account preference. Indicates that at the same time The learning materials browsed historically in the usage records of each youth account are substituted into the material dataset for query to obtain the category labels of the browsing preferences of each youth account.

[0021] Preferably, in step 5, the recommended corpus The calculation formula is as follows:

[0022] ,

[0023] In the formula, Indicates recommended corpus, Indicates that the data is grouped by preference The category labels that each youth account browses in the data are used to filter out the learning materials with the same category labels in the material data, which is the recommendation list. Represents the learning materials browsed historically in the account dataset usage record. It means to remove the learning materials that have been browsed in the past from the list of recommended materials and get the materials to be recommended. Indicates that the recommended materials will be combined The keywords that appear frequently in the language expression of each teenager are used to obtain recommended corpus for personalized learning.

[0024] Preferably, in step 6, if during the observation period If the number of browsing times, browsing duration, and browsing quantity of learning materials in the account dataset usage record increases, it means that teenagers are highly interested in personalized learning materials and continue to follow the recommended corpus. Recommended lists in the Recommended list recommend personalized learning materials to youth accounts.

[0025] Preferably, in step 6, if during the observation period If the number of browsing times, browsing duration, and browsing quantity of learning materials in the account dataset usage record do not increase, it means that teenagers have low interest in personalized learning materials. The corpus analysis module reanalyzes and generates a preference dataset. and recommended corpus , re-recommend personalized learning materials to youth accounts.

[0026] A personalized education recommendation system based on a youth corpus, including a data collection module and a corpus analysis module;

[0027] The data collection module includes a hot data unit, an account data unit and a material data unit. The hot data unit collects hot data sets in real time through a network connection to a big data platform. The hot data sets include language keywords that appear frequently in the process of teenagers' language expression. The account data set collects account data sets through a network connection to an education platform. The account data set includes usage records of all teenagers' accounts. The material data unit collects material data sets through a network connection to an education platform. The material data sets include all material data used to recommend teenagers for learning. The data collection module transmits the hot data sets, account data sets and material data sets to the corpus analysis module through the network.

[0028] The corpus analysis module includes a preference analysis unit, a personality recommendation unit and a monitoring management unit. The preference analysis unit generates a preference data set according to the account data set and the material data set. and transmitted to the personality recommendation unit through the network, and the personality recommendation unit is based on the preference data group , popular data sets and material data sets, analyze and generate recommended corpora , and transmitted to the monitoring management unit through the network, and the monitoring management unit Recommend personalized learning materials to each youth account, and the monitoring management unit is set with a fixed-length observation cycle , and monitor the observation cycle Based on the changes in the internal account data set, we can determine the degree of interest of teenagers in personalized learning materials and then adjust the personalized recommendation strategy accordingly.

[0029] Compared with the prior art, the present invention provides a personalized education recommendation method and system based on a youth corpus, which has the following beneficial effects:

[0030] 1. The present invention uses a data collection module to comprehensively collect popular keywords in the current time point in the network language, individual differences of the youth group, and all learning materials suitable for the current cognitive stage of the youth, and classify them into popular data sets, account data sets, and material data sets, which helps to improve the relevance and effectiveness of subsequent personalized education recommendations. The corpus analysis module analyzes and generates a preference data set based on the account data set and the material data set. , get the category labels of each youth account’s browsing preferences, and personalized education recommendations for better learning results.

[0031] 2. The present invention uses a corpus analysis module to analyze the preference data group , popular data sets and material data sets, analyze and generate recommended corpora , get recommended corpus for personalized learning, the extensive use of popular keywords helps young people maintain a high degree of sensitivity and interest in the language learning process, and improves their ability to adapt to and understand language changes by constantly contacting and learning recommended corpus and learning materials of interest. The corpus analysis module Recommend personalized learning materials to each youth account, and set a fixed observation period for the corpus analysis module , and monitor the observation cycle Based on the changes in the internal account data set, we can determine the degree of interest of teenagers in personalized learning materials, and then adjust the personalized recommendation strategy accordingly, flexibly monitor the changes in preferences, and optimize and update them more timely. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 It is a step diagram of the method of the present invention;

[0033] Figure 2 It is a schematic diagram of the system flow of the present invention. DETAILED DESCRIPTION

[0034] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0035] The traditional personalized education recommendation method based on the adolescent corpus ignores the differences between adolescents and adopts a unified recommendation strategy, which results in poor educational learning effects. In addition, in actual use, there is a lack of an effective feedback mechanism to collect adolescents' feedback on the recommended content, and it is impossible to optimize and improve the recommendation strategy in a timely manner. Therefore, a personalized education recommendation method and system based on the adolescent corpus is provided. Please refer to Figure 1-Figure 2 , the method comprises the following steps:

[0036] Step 1: The data collection module connects to the big data platform through the network to obtain the language data that appears frequently in the process of teenagers' language expression in real time, and uses natural language processing (NLP) technology to pre-process the collected language data, and then extracts keywords from the pre-processed data to form a popular data set. The big data platform includes but is not limited to online channels such as social software, news media, online forums, and school forums. The expression of the popular data set is , to Corresponding to each high-frequency keyword in the language expression of teenagers, Indicates the time point when the keyword appears frequently, from 1 to The keywords that frequently appear in teenagers’ language expression are Collect rich and diverse popular keywords, which is conducive to the subsequent matching of learning materials with relevant interests, and personalized recommendation of educational content, which helps to enhance the language expression and social skills of young people;

[0037] Step 2: The data collection module obtains the usage records of youth accounts through the network connection education platform, and forms the usage records of all youth accounts into an account data set. The expression of the account data set is: , to The usage records of each youth account are recorded in turn, including the learning materials browsed in the past and the number of times and duration of browsing of each learning material. Indicates the number of browsing of learning materials in the usage record. Learning materials include but are not limited to books, movies, videos, and interactive games. 1 to Indicates the number of youth accounts Collect the learning materials browsed by each youth account, which is conducive to the subsequent targeted analysis of the browsing preferences of each youth user;

[0038] Step 3: The data collection module obtains all the material data used to recommend learning to teenagers through the network connection education platform, and forms all the material data used to recommend learning into a material data set. The expression of the material data set is: , to The corresponding data of each material used to recommend learning to young people are listed in turn. The material data includes learning content, learning form and material evaluation. The material evaluation directly reflects the feedback of young people on the recommended content, such as satisfaction rating and comments. Indicates the type label of the learning materials. All learning materials are set with corresponding type labels. The type labels are divided into five major categories and 22 basic categories according to the unified standards in the "Chinese Library Classification System". Indicates the number of learning materials share;

[0039] Step 4: The corpus analysis module generates a preference data set based on the account data set and the material data set. , and its calculation formula is as follows:

[0040] ,

[0041] In the formula, represents the preference data set, Represents all the material data in the comprehensive material dataset used to recommend learning to teenagers, and is marked as , Indicates the account data set The usage records of youth accounts, Indicates that the learning materials browsed in the usage record are substituted into the material dataset for query, and the first Category tags that teenagers browse by account preference. Indicates that at the same time The learning materials browsed in the history of each youth account usage record are substituted into the material data set for query, and the category labels of each youth account's browsing preferences are obtained, so that personalized education recommendation learning effects are better;

[0042] Step 5: Corpus analysis module based on preference data group , popular data sets and material data sets, analyze and generate recommended corpora , and its calculation formula is as follows:

[0043] ,

[0044] In the formula, Indicates recommended corpus, Indicates that the data is grouped by preference The category labels that each youth account browses in the data are used to filter out the learning materials with the same category labels in the material data, which is the recommendation list. Represents the learning materials browsed historically in the account dataset usage record. It means to remove the learning materials that have been browsed in the past from the list of recommended materials to obtain the materials to be recommended, so as to avoid repeated recommendation of materials that the young users have already browsed and studied. Indicates that the recommended materials will be combined Keywords that frequently appear in the language expression of teenagers are used to obtain recommended corpus for personalized learning. The extensive use of popular keywords helps teenagers maintain a high degree of sensitivity and interest in the language learning process. By constantly contacting and learning recommended corpus and learning materials of interest, teenagers' ability to adapt to and understand language changes is improved;

[0045] Step 6: Corpus analysis module based on recommended corpus Recommend personalized learning materials to each youth account, and set a fixed observation period for the corpus analysis module , and monitor the observation cycle Based on the changes in the account data set, we can determine the interest of teenagers in personalized learning materials and adjust the personalized recommendation strategy accordingly. If the number of browsing times, browsing duration, and browsing quantity of learning materials in the account dataset usage record increases, it means that teenagers are highly interested in personalized learning materials and continue to follow the recommended corpus. Recommended list in the recommendation list recommends personalized learning materials to the youth account. If If the number of browsing times, browsing duration, and browsing quantity of learning materials in the account dataset usage record do not increase, it means that teenagers have low interest in personalized learning materials. The corpus analysis module reanalyzes and generates a preference dataset. and recommended corpus , re-recommend personalized learning materials to youth accounts, flexibly monitor preference changes and optimize updates more timely.

[0046] The above-mentioned personalized education recommendation method based on the youth corpus is realized by relying on the personalized education recommendation system based on the youth corpus, which includes a data collection module and a corpus analysis module;

[0047] The data collection module includes a hot data unit, an account data unit and a material data unit. The hot data unit collects hot data sets in real time through a network connection to a big data platform. The hot data sets include language keywords that appear frequently in the process of teenagers' language expression. The account data set collects account data sets through a network connection to an education platform. The account data set includes the usage records of all teenagers' accounts. The material data unit collects material data sets through a network connection to an education platform. The material data set includes all material data used to recommend teenagers to study. The data collection module transmits the hot data sets, account data sets and material data sets to the corpus analysis module through the network, comprehensively collects hot keywords in the current time point in the network language, individual differences of the teenager group and all learning materials suitable for the current cognitive stage of teenagers, which helps to improve the relevance and effectiveness of subsequent personalized education recommendations, meet the emotional and psychological needs of each student, and provide sufficient emotional support and motivation;

[0048] The corpus analysis module includes a preference analysis unit, a personality recommendation unit, and a monitoring management unit. The preference analysis unit generates a preference data set based on the account data set and the material data set. , and transmitted to the personalized recommendation unit through the network. The personalized education recommendation learning effect is better. The personalized recommendation unit is based on the preference data group. , popular data sets and material data sets, analyze and generate recommended corpora , and transmitted to the monitoring management unit through the network. The monitoring management unit recommends Recommend personalized learning materials to each youth account, and set a fixed observation period for the monitoring management unit , and monitor the observation cycle Based on the changes in the account data set, we can determine the interest of teenagers in personalized learning materials and adjust the personalized recommendation strategy accordingly. If the number of views, browsing time and number of views of learning materials in the account data set usage record do not increase, it means that teenagers have low interest in personalized learning materials, their browsing preferences have changed, and their cognitive stage has also been upgraded accordingly. It is necessary to recommend personalized learning materials to teenagers' accounts again, flexibly monitor preference changes, and optimize and update them more timely.

[0049] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A personalized education recommendation method based on a youth corpus, characterized by: The following steps are involved: Step 1: The data collection module connects to the big data platform through the network to obtain the language data that appears frequently in the language expression process of teenagers in real time, and uses natural language processing (NLP) technology to pre-process the collected language data, and then extracts keywords from the pre-processed data to form a popular data set; Step 2: The data collection module obtains usage records of youth accounts through a network connection to the education platform, and forms an account data set with usage records of all youth accounts; Step 3: The data collection module obtains all the material data for recommending young people to study through the Internet connection education platform, and forms all the material data for recommending study into a material data set; Step 4: The corpus analysis module generates a preference data set based on the account data set and the material data set. , get the browsing preferences of each youth account; Step 5: Corpus analysis module based on preference data group , popular data sets and material data sets, analyze and generate recommended corpora , get the recommended corpus for personalized learning; Step 6: Corpus analysis module based on recommended corpus Recommend personalized learning materials to each youth account, and set a fixed observation period for the corpus analysis module , and monitor the observation cycle Based on the changes in the internal account data set, we can determine the interest of teenagers in personalized learning materials and adjust the personalized recommendation strategy accordingly. In step 4, the preference data set The calculation formula is as follows: ， In the formula, represents the preference data set, Represents all the material data in the comprehensive material dataset used to recommend learning to teenagers, and is marked as , Indicates the account data set The usage records of youth accounts, Indicates that the learning materials browsed in the usage record are substituted into the material dataset for query, and the first Category tags that teenagers browse by account preference. Indicates that at the same time Substitute the historically browsed learning materials in the usage records of each youth account into the material data set for query, and obtain the category label of each youth account's browsing preference; In step 5, the recommended corpus The calculation formula is as follows: ， In the formula, Indicates recommended corpus, Indicates that the data is grouped by preference The category labels that each youth account browses in the data are used to filter out the learning materials with the same category labels in the material data, which is the recommendation list. Represents the learning materials browsed historically in the account dataset usage record. It means to remove the learning materials that have been browsed in the past from the list of recommended materials and get the materials to be recommended. Indicates that the recommended materials will be combined The keywords that appear frequently in the language expression of each teenager are used to obtain recommended corpus for personalized learning.

2. The personalized education recommendation method based on a youth corpus according to claim 1, characterized in that: In the step 1, the preprocessing process includes word segmentation, part-of-speech tagging and syntactic analysis. The word segmentation divides the language data that appears frequently into independent words. The part-of-speech tagging marks the grammatical attributes of the divided words. The syntactic analysis analyzes the structural relationship and overall semantics between each word based on the grammatical attribute markings, and selects keywords based on the overall semantics of the language data.

3. The personalized education recommendation method based on a youth corpus according to claim 2 is characterized by: In step 1, the expression of the popular data set is: , to Corresponding to each high-frequency keyword in the language expression of teenagers, Indicates the time point when the keyword appears frequently, from 1 to The keywords that frequently appear in teenagers’ language expression are indivual.

4. The personalized education recommendation method based on a youth corpus according to claim 3 is characterized by: In step 2, the expression of the account data set is: , to The usage records of each youth account are recorded in turn, including the learning materials browsed in the past and the number of times and duration of browsing of each learning material. Indicates the number of times the learning material is viewed in the usage record, from 1 to Indicates the number of youth accounts indivual.

5. The personalized education recommendation method based on a youth corpus according to claim 4 is characterized by: In step 3, the expression of the material data set is: , to Corresponding to each material data used to recommend learning to young people, the material data includes learning content, learning form and material evaluation, Indicates the type of learning material, ranging from 1 to Indicates the number of learning materials share.

6. The personalized education recommendation method based on a youth corpus according to claim 5 is characterized by: In step 6, if during the observation period If the number of browsing times, browsing duration, and browsing quantity of learning materials in the account dataset usage record increases, it means that teenagers are highly interested in personalized learning materials and continue to follow the recommended corpus. Recommended lists in the Recommended list recommend personalized learning materials to youth accounts.

7. The personalized education recommendation method based on a youth corpus according to claim 6 is characterized by: In step 6, if during the observation period If the number of browsing times, browsing duration, and browsing quantity of learning materials in the account dataset usage record do not increase, it means that teenagers have low interest in personalized learning materials. The corpus analysis module reanalyzes and generates a preference dataset. and recommended corpus , re-recommend personalized learning materials to youth accounts.

8. A personalized education recommendation system based on a youth corpus, applied to the personalized education recommendation method based on a youth corpus according to any one of claims 1 to 7, characterized in that: Including data collection module and corpus analysis module; The data collection module includes a hot data unit, an account data unit and a material data unit. The hot data unit collects hot data sets in real time through a network connection to a big data platform. The hot data sets include language keywords that appear frequently in the process of teenagers' language expression. The account data set collects account data sets through a network connection to an education platform. The account data set includes usage records of all teenagers' accounts. The material data unit collects material data sets through a network connection to an education platform. The material data sets include all material data used to recommend teenagers for learning. The data collection module transmits the hot data sets, account data sets and material data sets to the corpus analysis module through the network. The corpus analysis module includes a preference analysis unit, a personality recommendation unit and a monitoring management unit. The preference analysis unit generates a preference data set according to the account data set and the material data set. and transmitted to the personality recommendation unit through the network, and the personality recommendation unit is based on the preference data group , popular data sets and material data sets, analyze and generate recommended corpora , and transmitted to the monitoring management unit through the network, and the monitoring management unit Recommend personalized learning materials to each youth account, and the monitoring management unit is set with a fixed-length observation cycle , and monitor the observation cycle Based on the changes in the internal account data set, we can determine the degree of interest of teenagers in personalized learning materials and then adjust the personalized recommendation strategy accordingly.

Images