Teaching resource personalized recommendation system and method based on deep learning

Abstract

The invention relates to the technical field of teaching resource recommendation, and discloses a teaching resource personalized recommendation system and method based on deep learning. The system comprises a student data acquisition module, a student data evaluation module and a medical education resource recommendation module. The student data acquisition module comprises a student idle time acquisition unit, a student professional data acquisition unit, a student assessment score acquisition unit and a student tendency data acquisition unit. According to the system, through collecting student professional names and previous assessment results uploaded by students, and tendency data of the students for each kind of educational resources, a lifting coefficient and a learning strategy coefficient are calculated respectively; and then, based on the lifting coefficient and the learning strategy coefficient, carrying out preliminary screening of the educational resources for each student in each rated idle time, and after the preliminary screening is completed, outputting a final recommendation result in combination with tendency data of the student for each educational resource. Therefore, the medical teaching resources can be recommended in a personalized manner in combination with the current learning defect of the user.

Description

Technical Field

[0001] The present invention relates to the technical field of teaching resource recommendation, and specifically to a personalized teaching resource recommendation system and method based on deep learning. Background Art

[0002] With the development of mobile Internet, there are various channels and methods to obtain information, and they are no longer limited by time and space. In the face of massive data resources, it is particularly important to efficiently obtain the required information, or even obtain other information that is more suitable than actively searching by yourself.

[0003] At present, various medical education apps are usually based on single courses and knowledge points, each forming its own knowledge barrier to consolidate its market position. The core demand of the online education industry is to better meet the needs of users for fragmented learning. Most existing platforms only provide a simple search function and cannot recommend learning resources to users. Users generally need to actively search for the resources they want to learn through search engines. Usually, there are many types of resources returned, and users need to filter and distinguish them, which cannot meet the personalized needs of users. Summary of the invention

[0004] In view of the deficiencies in the prior art, the present invention provides a personalized recommendation system and method for teaching resources based on deep learning, which has the advantages of personalized recommendation, etc., to solve the above technical problems.

[0005] To achieve the above-mentioned purpose, the present invention provides the following technical solutions: a personalized recommendation system for teaching resources based on deep learning, comprising a student data collection module, a student data evaluation module and a medical education resource recommendation module; The student data collection module includes a student free time collection unit, a student professional data collection unit, a student assessment score collection unit and a student tendency data collection unit. The student free time collection unit is used to collect the planning table actively uploaded by the student and automatically calculate the rated free time. The student professional data collection unit is used to collect the student's professional name uploaded by the student. The student assessment score collection unit is used to collect the student's last assessment score. The student tendency data collection unit is used to collect the student's tendency data for each educational resource and send it to the medical education resource recommendation module. The student data collection module sends the calculated rated free time, the student's professional name and the student's last assessment score to the student data evaluation module. The student data evaluation module calculates the learning strategy coefficient of each rated free time of the student based on the rated free time of the student, and the student data evaluation module also generates an improvement coefficient based on the student's last assessment score, and the student data evaluation module sends the learning strategy coefficient and the improvement coefficient of each rated free time of the student to the medical education resource recommendation module; The medical education resource recommendation module performs a preliminary screening of education resources for each student's rated free time based on the student's learning strategy coefficient and improvement coefficient, and outputs the final recommendation result in combination with the student's preference data for each education resource.

[0006] As a preferred technical solution of the present invention, the student free time collection unit collects the planning table actively uploaded by the student, and automatically calculates the specific expression of the rated free time as follows: ; in, represents the rated idle time dataset, They represent the rated idle time of the first section, ..., the Rated idle time, ..., Rated idle time, and Rated idle time The specific expression is as follows: ; in, Indicates The starting time of the segment planning, Indicates The planned end time of the segment.

[0007] As a preferred technical solution of the present invention, the specific expression of the student's last assessment score collected by the student assessment score collection unit is as follows: ; in, represents the assessment score dataset, They represent the first subject assessment results,..., The results of various subjects,... Various subject assessment results.

[0008] As a preferred technical solution of the present invention, the student preference data collection unit collects the student's preference data for each educational resource, specifically including the student's click rate, learning time, review times and play ratio for each educational resource, and its expression is as follows: ; in, represents the tendency dataset of educational resources, They represent the click rate of the first educational resource,..., The click rate of educational resources,... The click-through rate of educational resources, They represent the learning time of the first educational resource,..., The learning time of educational resources,... The learning time of educational resources Respectively represent the number of times the first educational resource is reviewed, ..., The number of times you review the educational resource, ... Number of times the educational resource is reviewed, They represent the playback ratio of the first educational resource, ..., The proportion of educational resources played,... The playback ratio of educational resources.

[0009] As a preferred technical solution of the present invention, the Click-through rate of educational resources The specific calculation expression is as follows: ; in, Indicates The number of clicks on educational resources, Indicates the number of clicks on all educational resources; The said The proportion of educational resources played The specific calculation expression is as follows: ; in, Indicates the total playback time. Indicates the elapsed playback time.

[0010] As a preferred technical solution of the present invention, the specific expression of the student data evaluation module calculating the learning strategy coefficient of each rated idle time of the student based on the rated idle time of the student is as follows: ; in, Indicates Segment learning strategy coefficient, represents a natural constant, Express Round up, Represents the adaptive coefficient, which is used to To round up, Indicates Rated idle time.

[0011] As a preferred technical solution of the present invention, the student data evaluation module also generates a specific expression of the improvement coefficient based on the student's last assessment score as follows: ; in, Indicates the maximum score. represents the lifting factor, Indicates Subject examination results, Express The scores of various subjects are summed up.

[0012] As a preferred technical solution of the present invention, the specific steps of the medical education resource recommendation module for preliminary screening of education resources for each student's rated free time based on the student's learning strategy coefficient and improvement coefficient are as follows: Step A1: Calculate the first Segment duration evaluation coefficient , the specific expression is: ; in, Indicates Segment learning strategy coefficient, represents the lifting factor; Step A2: Select the teaching resource library related to the first Segment duration evaluation coefficient All teaching resources with a difference value of less than 10% were taken as the preliminary screening results.

[0013] As a preferred technical solution of the present invention, the specific steps of the medical education resource recommendation module outputting the final recommendation result in combination with the trainees' preference data for each type of education resource are as follows: Step B1: Calculate the first The tendency coefficient of educational resources , its specific expression is as follows: ; in, Indicates The tendency coefficient of educational resources, Indicates The click-through rate of educational resources, Indicates The learning time of educational resources Indicates The number of times the educational resource is reviewed, Indicates The proportion of educational resources played; Step B2: Output and select the educational resource with the smallest propensity coefficient among all educational resources and recommend it to the user.

[0014] The present invention also provides a personalized recommendation method for teaching resources based on deep learning, based on the above-mentioned personalized recommendation system for teaching resources based on deep learning, comprising the following steps: Step 1: Collect the planning table uploaded by the students, automatically calculate the rated free time, collect the name of the students' majors and the last assessment results uploaded by the students, and collect the students' preference data for each educational resource; Step 2: Calculate the learning strategy coefficient of each rated free time of the student based on the rated free time of the student; Step 3: Generate an improvement coefficient based on the student’s last assessment result; Step 4: Conduct a preliminary screening of educational resources for each student’s rated free time based on the student’s learning strategy coefficient and improvement coefficient; Step 5: After the initial screening is completed, the final recommendation results are output based on the students' preference data for each educational resource.

[0015] Compared with the prior art, the present invention provides a personalized recommendation system and method for teaching resources based on deep learning, which has the following beneficial effects: The present invention collects the students' major names and last examination results uploaded by the students, and the students' inclination data for each educational resource, and calculates the improvement coefficient and the learning strategy coefficient respectively. Then, based on the improvement coefficient and the learning strategy coefficient, a preliminary screening of educational resources is performed for each student's each rated free time. After the preliminary screening is completed, the final recommendation result is output in combination with the students' inclination data for each educational resource, so that personalized recommendations for medical teaching resources can be made in combination with the user's current learning deficiencies. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a schematic diagram of the system framework of the present invention; Figure 2 It is a schematic diagram of the process of the present invention. DETAILED DESCRIPTION

[0017] 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.

[0018] See also Figure 1-Figure 2, a personalized recommendation system for teaching resources based on deep learning, including student data collection module, student data evaluation module and medical education resource recommendation module; The student data collection module includes a student free time collection unit, a student professional data collection unit, a student assessment score collection unit and a student tendency data collection unit. The student free time collection unit is used to collect the planning table actively uploaded by the students and automatically calculate the rated free time. Here, the scheduling data can be formatted into an image-like form through a deep learning algorithm. The schedule of each student or the course schedule of each student can be regarded as a two-dimensional matrix, and the corresponding free time can be obtained. The students here include employees or students. The student free time collection unit collects the planning table actively uploaded by the students and automatically calculates the specific expression of the rated free time as follows: ; in, represents the rated idle time dataset, They represent the rated idle time of the first section, ..., the Rated idle time, ..., Rated idle time, and Rated idle time The specific expression is as follows: ; in, Indicates The starting time of the segment planning, Indicates The planned end time of the segment, such as The planned time for this section is 15:00-19:00. The time period is 21:00-21:30, 21:00-19:00 = 2h. The student major data collection unit is used to collect the student major name uploaded by the student. The student assessment score collection unit is used to collect the student's last assessment score. The specific expression of the student assessment score collection unit collecting the student's last assessment score is as follows: ; in, represents the assessment score dataset, They represent the first subject assessment results,..., The results of various subjects,... The student preference data collection unit is used to collect the student's preference data for each educational resource. The student preference data collection unit collects the student's preference data for each educational resource, specifically including the student's click rate, study time, review times and playback ratio for each educational resource. The expression is as follows: ; in, represents the tendency dataset of educational resources, They represent the click rate of the first educational resource,..., The click rate of educational resources,... The click-through rate of educational resources, They represent the learning time of the first educational resource,..., The learning time of educational resources,... The learning time of educational resources Respectively represent the number of times the first educational resource is reviewed, ..., The number of times you review the educational resource, ... Number of times the educational resource is reviewed, They represent the playback ratio of the first educational resource, ..., The proportion of educational resources played,... The proportion of educational resources played Click-through rate of educational resources The specific calculation expression is as follows: ; in, Indicates The number of clicks on educational resources, Indicates the number of clicks on all educational resources; No. The proportion of educational resources played The specific calculation expression is as follows: ; in, Indicates the total playback time. Indicates the playing time and sends it to the medical education resource recommendation module. The student data collection module sends the calculated rated free time, the student's major name and the student's last assessment score to the student data evaluation module; The student data evaluation module calculates the learning strategy coefficient of each rated free time of the student based on the rated free time of the student. The student data evaluation module also generates an improvement coefficient based on the student's last assessment score. The student data evaluation module sends the learning strategy coefficient and the improvement coefficient of each rated free time of the student to the medical education resource recommendation module. The specific expression of the learning strategy coefficient of each rated free time of the student calculated by the student data evaluation module based on the rated free time of the student is as follows: ; in, Indicates Segment learning strategy coefficient, represents a natural constant, Express Round up, Represents the adaptive coefficient, which is used to To round, for example: , , Indicates The specific expression of the improvement coefficient generated by the student data evaluation module based on the student's last assessment score is as follows: ; in, Indicates the maximum score. represents the lifting factor, Indicates Subject examination results, Express The scores of various subjects are summed up; The medical education resource recommendation module performs preliminary screening of education resources for each student's rated free time based on the student's learning strategy coefficient and the improvement coefficient, and outputs the final recommendation results based on the student's preference data for each type of education resource. The specific steps for the medical education resource recommendation module to perform preliminary screening of education resources for each student's rated free time based on the student's learning strategy coefficient and the improvement coefficient are as follows: Step A1: Calculate the first Segment duration evaluation coefficient , the specific expression is: ; in, Indicates Segment learning strategy coefficient, Indicates the improvement coefficient. The higher the improvement coefficient, the greater the user's defect is, so the defect can be identified. Step A2: Select the teaching resource library related to the first Segment duration evaluation coefficient All teaching resources with a difference value of less than 10% were taken as the preliminary screening results; The specific steps for the medical education resource recommendation module to output the final recommendation results based on the students' preference data for each type of education resource are as follows: Step B1: Calculate the first The tendency coefficient of educational resources , its specific expression is as follows: ; in, Indicates The tendency coefficient of educational resources, Indicates The click-through rate of educational resources, Indicates The learning time of educational resources Indicates The number of times the educational resource is reviewed, Indicates The proportion of educational resources played, which reflects which resource users are least inclined to learn. The larger the value of is, the more it reflects the user's higher learning tendency towards this kind of educational resource, thus recommending the user's weak points to the user; Step B2: Output and select the educational resource with the smallest propensity coefficient among all educational resources and recommend it to the user.

[0019] The present invention also provides a personalized recommendation method for teaching resources based on deep learning, based on the above-mentioned personalized recommendation system for teaching resources based on deep learning, comprising the following steps: Step 1: Collect the planning table uploaded by the students, automatically calculate the rated free time, collect the name of the students' majors and the last assessment results uploaded by the students, and collect the students' preference data for each educational resource; Step 2: Calculate the learning strategy coefficient of each rated free time of the student based on the rated free time of the student; Step 3: Generate an improvement coefficient based on the student’s last assessment result; Step 4: Conduct a preliminary screening of educational resources for each student’s rated free time based on the student’s learning strategy coefficient and improvement coefficient; Step 5: After the initial screening is completed, the final recommendation results are output based on the students' preference data for each educational resource.

[0020] 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 recommendation system for teaching resources based on deep learning, characterized by: It includes student data collection module, student data evaluation module and medical education resource recommendation module; The student data collection module includes a student free time collection unit, a student professional data collection unit, a student assessment score collection unit and a student tendency data collection unit. The student free time collection unit is used to collect the planning table actively uploaded by the student and automatically calculate the rated free time. The student professional data collection unit is used to collect the student's professional name uploaded by the student. The student assessment score collection unit is used to collect the student's last assessment score. The student tendency data collection unit is used to collect the student's tendency data for each educational resource and send it to the medical education resource recommendation module. The student data collection module sends the calculated rated free time, the student's professional name and the student's last assessment score to the student data evaluation module. The student data evaluation module calculates the learning strategy coefficient of each rated free time of the student based on the rated free time of the student, and the student data evaluation module also generates an improvement coefficient based on the student's last assessment score, and the student data evaluation module sends the learning strategy coefficient and the improvement coefficient of each rated free time of the student to the medical education resource recommendation module; The medical education resource recommendation module performs a preliminary screening of education resources for each student's rated free time based on the student's learning strategy coefficient and improvement coefficient, and outputs the final recommendation result in combination with the student's preference data for each education resource.

2. The deep learning-based personalized recommendation system for teaching resources according to claim 1, characterized in that: The student idle time collection unit collects the planning table actively uploaded by the student and automatically calculates the specific expression of the rated idle time as follows: ； in, represents the rated idle time dataset, They represent the rated idle time of the first section, ..., the Rated idle time, ..., Rated idle time, and Rated idle time The specific expression is as follows: ； in, Indicates The starting time of the segment planning, Indicates The planned end time of the segment.

3. The personalized recommendation system for teaching resources based on deep learning according to claim 2 is characterized in that: The specific expression for the student assessment score collection unit to collect the student's last assessment score is as follows: ； in, represents the assessment score dataset, They represent the first subject assessment results,..., The results of various subjects,... Various subject assessment results.

4. The personalized recommendation system for teaching resources based on deep learning according to claim 3 is characterized by: The student preference data collection unit collects the student preference data for each educational resource, specifically including the student's click rate, learning time, review times and play ratio for each educational resource, and the expression is as follows: ； in, represents the tendency dataset of educational resources, They represent the click rate of the first educational resource,..., The click rate of educational resources,... The click-through rate of educational resources, They represent the learning time of the first educational resource,..., The learning time of educational resources,... The learning time of educational resources Respectively represent the number of times the first educational resource is reviewed, ..., The number of times you review the educational resource, ... The number of times the educational resource is reviewed, They represent the playback ratio of the first educational resource, ..., The proportion of educational resources played,... The playback ratio of educational resources.

5. The personalized recommendation system for teaching resources based on deep learning according to claim 4 is characterized by: The said Click-through rate of educational resources The specific calculation expression is as follows: ； in, Indicates The number of clicks on educational resources, Indicates the number of clicks on all educational resources; The said The proportion of educational resources played The specific calculation expression is as follows: ； in, Indicates the total playback time. Indicates the elapsed playback time.

6. The personalized recommendation system for teaching resources based on deep learning according to claim 5, characterized in that: The specific expression for calculating the learning strategy coefficient of each rated free time of the student based on the rated free time of the student by the student data evaluation module is as follows: ； in, Indicates Segment learning strategy coefficient, represents a natural constant, Express Round up, Represents the adaptive coefficient, which is used to To round up, Indicates Rated idle time.

7. The deep learning-based personalized recommendation system for teaching resources according to claim 6, characterized in that: The student data evaluation module also generates a specific expression for the improvement coefficient based on the student's last assessment score as follows: ； in, Indicates the maximum score. represents the lifting factor, Indicates Subject examination results, Express The scores of various subjects are summed up.

8. The deep learning-based personalized recommendation system for teaching resources according to claim 7, characterized in that: The specific steps of the medical education resource recommendation module for preliminary screening of education resources for each student's rated free time based on the student's learning strategy coefficient and improvement coefficient are as follows: Step A1: Calculate the first Segment duration evaluation coefficient , the specific expression is: ； in, Indicates Segment learning strategy coefficient, represents the lifting factor; Step A2: Select the teaching resource library related to the first Segment duration evaluation coefficient All teaching resources with a difference value of less than 10% were taken as the preliminary screening results.

9. The deep learning-based personalized recommendation system for teaching resources according to claim 8, characterized in that: The specific steps of the medical education resource recommendation module outputting the final recommendation result in combination with the trainees' preference data for each type of education resource are as follows: Step B1: Calculate the first The tendency coefficient of educational resources , its specific expression is as follows: ； in, Indicates The tendency coefficient of educational resources, Indicates The click-through rate of educational resources, Indicates The learning time of educational resources Indicates The number of times the educational resource is reviewed, Indicates The proportion of educational resources played; Step B2: Output and select the educational resource with the smallest propensity coefficient among all educational resources and recommend it to the user.

10. A personalized recommendation method for teaching resources based on deep learning, a personalized recommendation system for teaching resources based on deep learning according to any one of claims 1 to 9, characterized in that: The following steps are involved: Step 1: Collect the planning table uploaded by the students, automatically calculate the rated free time, collect the name of the students' majors and the last assessment results uploaded by the students, and collect the students' preference data for each educational resource; Step 2: Calculate the learning strategy coefficient of each rated free time of the student based on the rated free time of the student; Step 3: Generate an improvement coefficient based on the student’s last assessment result; Step 4: Conduct a preliminary screening of educational resources for each student’s rated free time based on the student’s learning strategy coefficient and improvement coefficient; Step 5: After the initial screening is completed, the final recommendation results are output based on the students' preference data for each educational resource.

Images