阅读说明：本技术 一种基于大数据的个性化自适应网上学习方法 (Personalized self-adaptive online learning method based on big data ) 是由 张元平 于 2019-08-06 设计创作，主要内容包括：本发明提出的一种基于大数据的个性化自适应网上学习方法,包括：用户网上做题数据进行跟踪,并根据用户做题记录划分学习板块；根据用户做题数据,计算用户各学习板块的错题率,并根据各学习板块的错题率向用户推荐习题。本发明中,以单个用户为单位,实现了对每一个用户的针对性指导,有利于实现每一个用户的最佳最适宜的学习方法,从而通过因材施教,保证每一用户的学习效率和学习效果。(The invention provides a personalized self-adaptive online learning method based on big data, which comprises the following steps: tracking the data of the user doing questions on the internet, and dividing learning boards according to the records of the user doing questions; and calculating the wrong question rate of each learning plate of the user according to the question making data of the user, and recommending the exercises to the user according to the wrong question rate of each learning plate. In the invention, the single user is taken as a unit, the targeted guidance of each user is realized, and the optimal and most suitable learning method of each user is favorably realized, so that the learning efficiency and the learning effect of each user are ensured by the aid of the teaching according to the factors.)

1. A personalized self-adaptive online learning method based on big data is characterized by comprising the following steps:

tracking data of doing questions on the user network, and dividing learning boards according to user question making records;

and calculating the wrong question rate of each learning plate of the user according to the question making data of the user, and recommending the exercises to the user according to the wrong question rate of each learning plate.

2. The big data-based personalized adaptive online learning method according to claim 1, specifically comprising the steps of:

s1, taking each chapter of each subject as a minimum question bank, wherein each minimum question bank stores the exercises of the corresponding chapter, and the exercises in the minimum question bank are divided into a plurality of question sets according to the difficulty level;

s2, obtaining user question making data for evaluation, and counting the wrong question rate of each subject in the user question making data;

s3, comparing thresholds through the fault rate of each subject to obtain weak subjects of the user;

s4, counting the wrong question rate of each chapter in the user weak subject according to the user question data for evaluation;

s5, comparing thresholds through the wrong topic rate of each chapter to obtain a weak chapter in a user weak subject;

and S6, selecting exercise forming objects from the minimum item libraries according to the user weak items and the user weak sections in the user weak items, and recommending the exercise forming objects to the user.

3. The personalized adaptive online learning method based on big data according to claim 2, further comprising the steps of SA: setting a plurality of exercise object forms, wherein each exercise object form is provided with a corresponding question type composition proportion; in step S1, each question set is further divided into a plurality of question types according to the question form;

in step S6, a practice subject is selected from the minimum subject libraries in accordance with the form of the practice subject selected by the user, and recommended to the user.

4. The personalized adaptive online learning method based on big data as claimed in claim 3, wherein the exercise object forms include a paper pattern consisting of each chapter content and each topic type, a chapter exercise pattern consisting of a single chapter content and each topic type, and an emphasis exercise pattern consisting of a single chapter single topic type.

5. The personalized adaptive web learning method based on big data according to claim 3, further comprising the step SB: setting an average threshold value, and setting a plurality of wrong question rate threshold values corresponding to the difficulty degree in the minimum question bank;

in step S3, the subject whose error rate is greater than the average threshold is obtained as the weak subject of the user; in step S5, a chapter with a higher error rate than the average threshold is obtained as a weak chapter of the user;

step S6 specifically includes: and selecting exercises from the exercise sets corresponding to the error rate according to the exercise object form selected by the user and the error rate of each chapter in the subject to form an exercise object, and pushing the exercise object to the user.

6. The personalized and adaptive online learning method based on big data as claimed in claim 5, wherein in step S6, all the exercises in the newly generated exercise object are not existed in the user exercise data for evaluation.

7. The personalized adaptive online learning method based on big data as claimed in claim 2, wherein in step S2, the user question data for evaluation is the user question data in a preset evaluation period; or presetting the question data of the user with the question amount.

8. The method of claim 7, wherein the preset evaluation period and the preset number of questions are set by a user.

Technical Field

The invention relates to the technical field of online learning, in particular to a personalized self-adaptive online learning method based on big data.

Background

With the continuous innovation and rapid development of information technology, the network technology is updated day by day, so that the generated mass data is more and more emphasized, and the big data era has come. In recent years, data mining, data analysis, and data-based content recommendation systems have been widely used in various industries. In the field of education, how to enable students to easily learn, teachers can easily teach the students to become a problem that more and more people pay attention to. As the application of big data in the field of education, personalized adaptive learning becomes a typical application case.

Data mining work has been widely spread in recent years, but research work on online learning of personalized recommendations by learners has been insufficient. The existing learning effect evaluation methods are all simple grade division, and the methods applying more accurate big data analysis and prediction are less, so that the influence of the individual characteristics (such as learning ability, learning efficiency, sensitivity to the change of learning content and the like) of a learner on the learning effect is not fully considered.

Disclosure of Invention

Based on the technical problems in the background art, the invention provides a personalized self-adaptive online learning method based on big data.

The invention provides a personalized self-adaptive online learning method based on big data, which comprises the following steps:

tracking data of doing questions on the user network, and dividing learning boards according to user question making records;

and calculating the wrong question rate of each learning plate of the user according to the question making data of the user, and recommending the exercises to the user according to the wrong question rate of each learning plate.

Preferably, the method specifically comprises the following steps:

s1, taking each chapter of each subject as a minimum question bank, wherein each minimum question bank stores the exercises of the corresponding chapter, and the exercises in the minimum question bank are divided into a plurality of question sets according to the difficulty level;

s2, obtaining user question making data for evaluation, and counting the wrong question rate of each subject in the user question making data;

s3, comparing thresholds through the fault rate of each subject to obtain weak subjects of the user;

s4, counting the wrong question rate of each chapter in the user weak subject according to the user question data for evaluation;

s5, comparing thresholds through the wrong topic rate of each chapter to obtain a weak chapter in a user weak subject;

and S6, selecting exercise forming objects from the minimum item libraries according to the user weak items and the user weak sections in the user weak items, and recommending the exercise forming objects to the user.

Preferably, the method further comprises the following steps of SA: setting a plurality of exercise object forms, wherein each exercise object form is provided with a corresponding question type composition proportion; in step S1, each question set is further divided into a plurality of question types according to the question form;

in step S6, a practice subject is selected from the minimum subject libraries in accordance with the form of the practice subject selected by the user, and recommended to the user.

Preferably, the exercise target form includes a test paper pattern including each chapter content and each topic type, a chapter exercise pattern including a single chapter content and each topic type, and a focus exercise pattern including a single chapter single topic type.

Preferably, the method further comprises the step SB: setting an average threshold value, and setting a plurality of wrong question rate threshold values corresponding to the difficulty degree in the minimum question bank;

in step S3, the subject whose error rate is greater than the average threshold is obtained as the weak subject of the user; in step S5, a chapter with a higher error rate than the average threshold is obtained as a weak chapter of the user;

step S6 specifically includes: and selecting exercises from the exercise sets corresponding to the error rate according to the exercise object form selected by the user and the error rate of each chapter in the subject to form an exercise object, and pushing the exercise object to the user.

Preferably, in step S6, all of the exercise questions newly generated in the exercise target are not present in the user exercise data for evaluation.

Preferably, in step S2, the user question data used for evaluation is the question data of the user in the preset evaluation period; or presetting the question data of the user with the question amount.

Preferably, the preset evaluation period and the preset amount are set by a user.

According to the personalized self-adaptive online learning method based on the big data, the learning progress of the user is tracked in real time by summarizing the problem data, weak links in the learning of the user are summarized in real time by the problem error rate, and therefore problem recommendation is conducted on the weak links of the user. Therefore, the invention realizes the comprehensive understanding of the learning condition of the user according to the exercise data, indicates the learning direction to the user through exercise recommendation and provides an effective learning improvement means.

In addition, in the invention, the single user is taken as a unit, the targeted guidance of each user is realized, and the optimal and most suitable learning method of each user is facilitated, so that the learning efficiency and the learning effect of each user are ensured through the teaching according to the factors.

Drawings

Fig. 1 is a flow chart of a personalized adaptive online learning method based on big data according to the present invention.

Detailed Description

Referring to fig. 1, the personalized adaptive online learning method based on big data provided by the invention comprises the following steps:

and tracking the on-line problem-making data of the user, and dividing the learning board blocks according to the user problem-making records.

And calculating the wrong question rate of each learning plate of the user according to the question making data of the user, and recommending the exercises to the user according to the wrong question rate of each learning plate.

Therefore, in the embodiment, the learning progress of the user can be tracked in real time by summarizing the problem data, and weak links in the learning process of the user can be summarized in real time by the problem error rate, so that the problem recommendation can be performed on the weak links of the user. Therefore, the embodiment realizes comprehensive understanding of the learning condition of the user according to the exercise data, indicates the learning direction to the user through exercise recommendation, and provides an effective learning improvement means.

In addition, in the implementation, the single user is taken as a unit, the targeted guidance of each user is realized, the best and most suitable learning method of each user is facilitated, and therefore the learning efficiency and the learning effect of each user are guaranteed through the teaching of the factors.

The personalized self-adaptive online learning method based on big data in the embodiment specifically comprises the following steps:

s1, each chapter of each subject is used as a minimum subject library, each minimum subject library stores the problem of the corresponding chapter, and the problem in the minimum subject library is divided into a plurality of subject sets according to the difficulty degree.

In the step, the setting of the question sets lays a foundation for providing exercises with different difficulties according to the requirements of users.

And S2, acquiring user question making data for evaluation, and counting the question error rate of each subject in the user question making data.

Specifically, in this step, the user question data used for evaluation is the user question data in a preset evaluation period, for example, one month is used as one preset evaluation period, and then the question error rate of each subject is counted every three months according to the user question data in the last three months of the user, so that the user learning condition is evaluated and detected periodically in time. Specifically, the preset evaluation period may be set by the user autonomously, or set specifically according to the learning conditions of different users, for example, a user with many weak subjects and a high wrong subject rate may adopt a shorter preset evaluation period, for example, two weeks; users with few weak subjects and low wrong-subject rate can adopt a longer preset evaluation period, for example, two months. Therefore, for users with poor learning, the learning progress condition can be clearly mastered and timely adjusted by shortening the preset evaluation period; for the well-studied users, the redundant exercise can be avoided, and the study quality is improved.