CN112214688A

CN112214688A - Recommendation method and device for repair course selection and server

Info

Publication number: CN112214688A
Application number: CN202011083928.2A
Authority: CN
Inventors: 田冷; 王泽川; 黄诗慧; 黄灿; 王恒力; 柴晓龙
Original assignee: China University of Petroleum Beijing
Current assignee: China University of Petroleum Beijing
Priority date: 2020-10-12
Filing date: 2020-10-12
Publication date: 2021-01-12

Abstract

The specification provides a recommendation method, a recommendation device and a server for a repair course. Based on the method, the score data of the preset basic lesson which the target user has repaired can be retrieved and obtained according to the identity of the target user carried in the course selection and repair recommendation request initiated by the target user; calling a pre-trained preset election and repair course score prediction model to obtain a corresponding election and repair course score prediction result according to the score data of the basic course preset by the target user; and then, the result of the grading prediction of the selected course can be used as a basis, the selected course with relatively high learning potential of the target user is determined from a plurality of selected courses to be used as a matched selected course, and the target user is recommended. Therefore, the system can help the student user to efficiently and accurately find the selection and repair course matched with the current learning condition of the student user and suitable for the student user to learn, and the technical problem that the student user cannot efficiently and accurately determine the selection and repair course suitable for the student user is solved.

Description

Recommendation method and device for repair course selection and server

Technical Field

The specification belongs to the technical field of internet, and particularly relates to a method, a device and a server for recommending a repair course.

Background

In course learning of schools such as universities, students can relatively autonomously select some courses as their own optional courses after finishing some basic courses which need to be taken. When the students choose to repair the courses, the students are often in a poor state and do not know which courses are more suitable for learning.

Usually, students consult acquaintances or sisters and choose their choice course according to their advice. However, since the number of preschool or sisters that the student can consult is relatively limited, and the individual condition of the consulted preschool or sister may be greatly different from the student. In addition, there is often some personal subjective factor to the advice provided by the captain or sister. The students are difficult to accurately and efficiently find the course selection and repair matched with the current learning condition of the students through the mode, and the course selection and repair device is suitable for the students to select to learn.

In view of the above technical problems, no effective solution has been proposed at present.

Disclosure of Invention

The specification provides a recommendation method, device and server for a repair course, which are used for helping student users to efficiently and accurately find a repair course which is matched with the current learning condition of the student users and is suitable for learning of the student users, and the technical problem that the student users cannot efficiently and accurately determine the repair course suitable for the student users is solved.

An embodiment of the present specification provides a method for recommending a repair course, including:

receiving a course selection and repair recommendation request of a target user; the course selection and repair recommendation request at least carries an identity of a target user;

acquiring grade data of a basic course preset by the target user by retrieving a course grade database according to the identity of the target user;

calling a preset repair course score prediction model to obtain a corresponding repair course score prediction result according to the score data of the basic course preset by the target user;

according to the result of the grading prediction of the repair courses, determining a repair course matched with a target user as a target repair course;

and generating and sending recommendation information about the target repair course to the target user.

In one embodiment, the repair course recommendation request further carries a professional label of the target user.

In one embodiment, the method further comprises:

according to the professional label of the target user, screening out a preset repair course score prediction model matched with the professional label of the target user from a plurality of preset repair course score prediction models;

correspondingly, calling a preset lesson selection score prediction model matched with the professional label of the target user to obtain a corresponding lesson selection score prediction result according to the score data of the basic lesson preset by the target user.

In one embodiment, the preset repair course achievement prediction model is established according to the following method:

collecting the course score data of the current student user as sample data; the course achievement data of the current student user comprises preset basic course achievement and optional course achievement of the current student user;

establishing a plurality of sample data sets according to the sample data; wherein, the sample data in the same sample data set at least comprises a lesson selection score of a lesson selection with the same lesson identification;

and training the initial model based on the neural network by using the plurality of sample data sets to establish a preset repair course result prediction model.

In one embodiment, before training the initial neural network-based model with the plurality of sample data sets, respectively, the method further comprises:

determining the difference between the selected course score and the preset basic course score of the sample data in each sample data set;

and removing the sample data of which the difference between the selected course achievement and the preset basic course achievement in each sample data set is greater than a preset difference threshold.

In one embodiment, the initial model includes a plurality of initial network model structures, wherein each of the plurality of initial network model structures corresponds to a repair course.

In one embodiment, in the training of the initial neural network-based model using the plurality of sample data sets, respectively, the method further includes:

by controlling the numerical value of the dropout parameter, the influence of dirty data on model training is reduced, so that overfitting is avoided.

In one embodiment, determining the repair course matched with the target user according to the repair course result prediction result comprises:

and screening the repairing courses with the predicted scores larger than a preset score threshold value from the plurality of repairing courses according to the repairing course score prediction result to serve as the repairing courses matched with the target user.

An embodiment of the present specification further provides a device for recommending a repair course, including:

the system comprises a receiving module, a selecting and repairing course recommending module and a selecting and repairing course recommending module, wherein the receiving module is used for receiving a selecting and repairing course recommending request of a target user; the course selection and repair recommendation request at least carries an identity of a target user;

the acquisition module is used for acquiring the score data of a basic course preset by the target user by searching the course score database according to the identity of the target user;

the calling module is used for calling a preset repair course score prediction model to obtain a corresponding repair course score prediction result according to the score data of the basic course preset by the target user;

the determining module is used for determining the repair course matched with the target user as the target repair course according to the repair course score prediction result;

and the recommending module is used for generating and sending recommending information about the target repair course to the target user.

The embodiment of the specification further provides a server, which comprises a processor and a memory for storing processor executable instructions, wherein the processor executes the instructions to receive a course selection recommendation request of a target user; the course selection and repair recommendation request at least carries an identity of a target user; acquiring grade data of a basic course preset by the target user by retrieving a course grade database according to the identity of the target user; calling a preset repair course score prediction model to obtain a corresponding repair course score prediction result according to the score data of the basic course preset by the target user; according to the result of the grading prediction of the repair courses, determining a repair course matched with a target user as a target repair course; and generating and sending recommendation information about the target repair course to the target user.

According to the recommendation method, device and server for the repair course provided by the specification, firstly, score data of a basic course preset by a target user is retrieved and acquired according to an identity of the target user carried in a repair course recommendation request initiated by the target user; calling a pre-trained preset election and repair course score prediction model to obtain a corresponding election and repair course score prediction result according to the score data of the basic course preset by the target user; and then, the result of the prediction of the result of the repair course can be used as a screening basis, the repair course with relatively high learning potential of the target user is determined from a plurality of repair courses and used as a matched repair course, and targeted recommendation is carried out on the target user. Therefore, the system can help the student user to efficiently and accurately find the selection and repair course matched with the current learning condition of the student user and suitable for the student user to learn, and the technical problem that the student user cannot efficiently and accurately determine the selection and repair course suitable for the student user is solved.

Drawings

In order to more clearly illustrate the embodiments of the present specification, the drawings needed to be used in the embodiments will be briefly described below, and the drawings in the following description are only some of the embodiments described in the present specification, and it is obvious to those skilled in the art that other drawings can be obtained according to the drawings without any creative effort.

Fig. 1 is a schematic flow chart of a recommendation method for a repair course according to an embodiment of the present disclosure;

FIG. 2 is a schematic diagram of a server according to an embodiment of the present disclosure;

fig. 3 is a schematic structural diagram of a recommendation device for a repair course according to an embodiment of the present disclosure;

fig. 4 is a schematic diagram of necessary lesson achievements obtained by applying the recommendation method for a elective lesson provided in the embodiment of the present specification in a scenario example;

fig. 5 is a schematic diagram of a computer program code used in a method for recommending a repair course according to an embodiment of the present specification, in an example scenario;

fig. 6 is a schematic diagram of an embodiment of a recommendation method for a repair course provided by an embodiment of the present specification, in a scenario example;

fig. 7 is a schematic diagram of an embodiment of a recommendation method for a repair course, to which the embodiments of the present specification are applied, in an example scenario.

Detailed Description

In order to make those skilled in the art better understand the technical solutions in the present specification, the technical solutions in the embodiments of the present specification will be clearly and completely described below with reference to the drawings in the embodiments of the present specification, and it is obvious that the described embodiments are only a part of the embodiments of the present specification, and not all of the embodiments. All other embodiments obtained by a person skilled in the art based on the embodiments in the present specification without any inventive step should fall within the scope of protection of the present specification.

Considering that based on the existing method, students can only consult acquaintances or sisters of learning when selecting self-repair class. The students often have limited students to consult with their college and schooler. Moreover, the individual condition of the consulted preschool or sister may be greatly different from the student, and the preschool or sister does not know the current specific learning condition of the student, so that the recommended repair course of the preschool or sister is not actually suitable for the student.

Further, limited in time and effort, a captain or sister may not learn all of the options. Furthermore, the recommendations provided by the captain and sister often carry subjective emotional factors of the individual. The repair course suggestions provided by the scholars or sisters are often incomplete and inaccurate. Therefore, students can hardly find the selection course which is really matched with the current learning condition of the students and is suitable for the selected learning.

In view of the root cause of the above problems, the present specification considers that most students will uniformly learn some necessary basic courses, such as calculus, linear algebra, college physics, etc., first. Often, the basic courses are associated with a number of repair courses. For example, oil reservoir engineering such as professional workover courses requires the use of related knowledge in calculus, linear algebra, and college physics. Therefore, the achievement of some basic courses of the student to represent the current learning situation of the student can reflect the learning potential of the student for learning some optional courses in the future from the side.

Based on the thought, the achievement data of a large number of basic courses to and from the students and the achievement of the selected optional course can be obtained in advance to serve as sample data. Through the big data analysis of the large amount of sample data, the association relation between the preset basic course score and the repair course score of the student can be determined. Furthermore, model training can be performed by using the large amount of sample data, so that the model learns the association relationship, and a preset lesson selection result prediction model which can predict the results of different lessons selected by the student based on the results of the preset basic lessons currently repaired by the student can be established.

Therefore, when a student wants to consult a repair class suitable for the student, the score data of a preset basic class of the student can be acquired firstly; calling a pre-trained preset election and repair course score prediction model to obtain a corresponding election and repair course score prediction result according to the score data of the preset basic course; and then, the result of the prediction of the result of the repair course can be used as a basis, the repair course with relatively high learning potential of the student is determined from a plurality of repair courses and is used as a matched repair course, and the repair course is recommended to the student. Therefore, the system can effectively help students to efficiently and accurately find the selection and repair courses matched with the current learning condition of the students and suitable for the students to learn, and the technical problem that the students cannot efficiently and accurately determine the selection and repair courses suitable for the students is solved.

Based on the above thought, referring to fig. 1, the embodiment of the present specification provides a recommendation method for a repair course. The method can be applied to the server side. In particular implementations, the method may include the following.

S101: receiving a course selection and repair recommendation request of a target user; and the course selection and repair recommendation request at least carries an identity of a target user.

In one embodiment, the recommendation method for the repair course can be particularly applied to a server of a data processing system for serving students to assist the students in course selection.

In this embodiment, the server may specifically include a background server that is applied to a data processing system side and is capable of implementing functions such as data transmission and data processing. Specifically, the server may be, for example, an electronic device having data operation, storage function and network interaction function. Alternatively, the server may be a software program running in the electronic device and providing support for data processing, storage and network interaction. In the present embodiment, the number of servers is not particularly limited. The server may specifically be one server, or may also be several servers, or a server cluster formed by several servers.

In one embodiment, the target user may be a student user who has finished repairing a preset basic course and is ready to select a repair course. For example, a student who selects a professional course is prepared to complete the study of a major course.

In an embodiment, the target user may specifically initiate a corresponding lecture selection recommendation request to the server through the terminal device by performing a corresponding operation on the terminal device. Correspondingly, the server receives the course selection and repair recommendation request of the target user.

The course selection recommending request is used for requesting the server to determine the request data which is matched with the current learning condition of the target user and is suitable for the course learned by the target user to recommend from the courses selectable by the target user. The course selection and repair recommendation request at least carries the identity identification information of the target user.

In this embodiment, the terminal device may specifically include a front-end electronic device that is applied to a user side and can implement functions such as data acquisition and data transmission. Specifically, the terminal device may be, for example, a desktop computer, a tablet computer, a notebook computer, a smart phone, and the like. Alternatively, the terminal device may be a software application capable of running in the electronic device. For example, it may be some APP running on a cell phone, etc.

In an embodiment, the identity of the target user may be specifically understood as identification information corresponding to the target user. Specifically, the identification may be a study number, an identification number, a name, and the like of the target user.

S102: and acquiring the score data of the basic course preset by the target user by searching the course score database according to the identity of the target user.

In an embodiment, the preset basic lesson may be specifically understood as a required lesson which has been previously repaired by the target user and has a certain correlation with a subsequent selected lesson to be selected by the target user.

Specifically, the preset basic lessons may include: higher mathematics I, higher mathematics II, oil layer physics, seepage mechanics, oil reservoir engineering, oil recovery engineering, etc.

The repair course may be a professional repair course selectable by the target user, and may include: high reservoir engineering, high seepage mechanics, reservoir numerical simulation, an enhanced recovery method, complex structure well theory and technology, oil extraction engineering scheme design and the like. The selection and repair course can be an optional professional selection and repair course or a plurality of optional professional selection and repair courses.

Of course, the above listed preset basic courses and optional courses are only illustrative. During specific implementation, according to specific application scenarios and processing requirements, other types of courses can be introduced as preset basic courses and optional courses. The present specification is not limited to these.

In one embodiment, the lesson achievement database may store achievement data of preset basic lessons previously reviewed by each student user. Specifically, each preset basic course score stored in the course score database carries an identity of the corresponding student user.

In one embodiment, the server may respond to a lecture selection recommendation request of a target user, and retrieve the class achievement database according to the identity of the target user carried in the lecture selection recommendation request, so as to obtain achievement data of a basic class preset by the target user.

S103: and calling a preset selecting and repairing course score prediction model to obtain a corresponding selecting and repairing course score prediction result according to the score data of the basic course preset by the target user.

In an embodiment, the aforementioned preset lesson election result prediction model can be specifically understood as a prediction model established in advance by using a large amount of sample data to learn and train the potential relationship between the student's preset basic lesson result and the lesson election result. The score of each optional repair course in the optional repair courses which are not yet learned by the user can be predicted through the model according to the current learning condition of the user (namely, the score data of the preset basic courses which are already repaired currently).

In one embodiment, in implementation, the server may input preset basic course result data of the target user as a model input into the preset repair course result prediction model. And operating the preset selecting and repairing course score prediction model to obtain corresponding model output. And determining the predicted result of each repair course of the target user according to the output of the model, and taking the result as the corresponding repair course result prediction result.

S104: and determining the repair course matched with the target user as the target repair course according to the repair course result prediction result.

In an embodiment, the target repair course may be specifically understood as a repair course that is matched with the current learning situation of the target user and is suitable for the target user to learn the to-be-recommended repair course.

In an embodiment, the determining the repair course matched with the target user according to the repair course result prediction result may include the following steps: and screening the repairing courses with the predicted scores larger than a preset score threshold value from the plurality of repairing courses according to the repairing course score prediction result to serve as the repairing courses matched with the target user.

In this embodiment, in specific implementation, the server may determine, according to the result of predicting the result of the repair course, a predicted result of each repair course of the target user; then the predicted achievement of each repair course is compared with a preset achievement threshold value to obtain a corresponding comparison result; and screening the repairing course with the predicted result larger than a preset result threshold value according to the comparison result to serve as the target repairing course matched with the target user.

In one embodiment, the preset performance thresholds for different student users may be different. Specifically, the server may generate a preset achievement threshold corresponding to each student according to an average value of achievement data of a basic course preset by each student.

In an embodiment, the determining the repair course matched with the target user according to the repair course result prediction result may further include the following steps in specific implementation: and arranging a plurality of the repair courses according to the prediction results of the repair course grades from high to low, and acquiring a preset number of repair courses in the front of the sequence as repair courses matched with the target user. The preset number can be determined according to the number of the selected courses to be selected by the target user or the total credit of the selected courses to be selected.

S105: and generating and sending recommendation information about the target repair course to the target user.

In one embodiment, after determining the target repair course matched with the target user in the above manner, the server may generate recommendation information about the target repair course, and send the recommendation information to the terminal device. The terminal equipment displays the recommendation information to the target user so that the recommendation server helps the selected course screened by the target user. Therefore, the target user can efficiently and accurately determine the matching with the current learning condition of the target user according to the recommendation information, and is suitable for selecting and learning the selected course which is learned by the target user.

In an embodiment, the repair course recommendation request may further carry a professional label of a professional to which the target user belongs. The professional label may be specifically understood as identification information for indicating a professional to which the target user belongs.

In one embodiment, it is considered that the basic lessons to be repaired and the professional repair lessons that can be repaired are different from each other for students in different professions. Therefore, before specific implementation, different specialties can be distinguished, corresponding preset lesson selection result prediction models are trained respectively for the different specialties, and a plurality of preset lesson selection result prediction models are obtained. Wherein, each preset lesson selection score prediction model corresponds to a specialty.

In one embodiment, when implemented, the method may further include the following: and screening a preset repairing course score prediction model matched with the professional label of the target user from a plurality of preset repairing course score prediction models according to the professional label of the target user, and calling the preset repairing course score prediction model matched with the professional label of the target user to obtain a corresponding repairing course score prediction result according to score data of a basic course preset by the target user when the method is implemented specifically. Therefore, a more accurate result of predicting the result of the selected course can be obtained.

In an embodiment, the preset repair course achievement prediction model may be specifically established according to the following manner:

s1: collecting the course score data of the current student user as sample data; the course achievement data of the current student user comprises preset basic course achievement and optional course achievement of the current student user;

s2: establishing a plurality of sample data sets according to the sample data; wherein, the sample data in the same sample data set at least comprises a lesson selection score of a lesson selection with the same lesson identification;

s3: and training the initial model based on the neural network by using the plurality of sample data sets to establish a preset repair course result prediction model.

In one embodiment, the server may collect pre-set basic lesson achievements from the class achievements database and the lesson selection achievements as sample data. Each piece of sample data comprises a preset basic course score of the current student user and a repair course score of the current student user.

In an embodiment, each of the plurality of sample data sets corresponds to a repair course. Different sample data in the same sample data set at least comprises the result of the selecting course with the same course identification (namely the same selecting course).

Specifically, when the sample data set is established, taking the establishment of the current sample data set corresponding to the current course selection as an example, the course identifier of the course selection corresponding to the course selection score contained in each sample data can be searched first; and finding and dividing the sample data containing the course identification corresponding to the current selection course into the current sample data set. The sample data in the sample data set at least contains the result of the current selection course.

In addition, the sample data in the current sample data set may be preprocessed, for example, the result of the repair course of other repair courses except the current repair course in the sample data may be deleted.

According to the mode, a plurality of sample data sets respectively corresponding to the selected courses can be established and obtained.

In one embodiment, during specific training, an initial model based on a neural network (e.g., ANN or CNN, etc.) may be constructed; and then, the initial model can be trained by utilizing the plurality of sample data sets respectively to obtain a preset lesson selection result prediction model.

In one embodiment, the initial model may include a plurality of initial network model structures. Wherein each initial network model structure corresponds to a repair course.

Correspondingly, the initial model is trained by using a plurality of sample data sets, and the training may include the following steps: and respectively training the initial network model structures corresponding to the initial models by using the sample data sets corresponding to the selected courses, so that the network structures can respectively learn the association relationship between the corresponding selected course scores and the preset basic course scores.

In an embodiment, before the training of the initial model based on the neural network by using the plurality of sample data sets, the method may further include the following steps: determining the difference between the selected course score and the preset basic course score of the sample data in each sample data set; and removing the sample data of which the difference between the selected course achievement and the preset basic course achievement in each sample data set is greater than a preset difference threshold.

In the embodiment, the reference value is low in consideration that some sample data can have great contingency. For example, one sample data set corresponding to the course a is: the score of a preset basic course 1 of the student A is 92, the score of a preset basic course 2 is 95, the score of a preset basic course 3 is 89, and the score of a preset basic course 4 is 91; the score of the first selection course A of the student is 32.

In the sample data, the difference degree between the preset basic class achievement and the repair class achievement of the student A is unreasonable. This is a high probability due to the individual factors of the student nail. For example, it may be that the student A does not learn to choose course A carefully and wastes his or her potential. Is a relatively occasional situation. Therefore, such sample data does not reflect the association relationship between the pre-set basic lesson achievement and the repair lesson A achievement which are commonly existed in most student groups under normal conditions. If the sample data is used in the subsequent model training process, the accuracy of the established repair course performance prediction model is influenced due to the introduction of accidental errors.

In one embodiment, in order to reduce the introduction of accidental errors and obtain sample data with relatively better effect and more universality, under the condition that the acquired result data of the set basic course result and/or the optional repair course result of the student user has unqualified results, the result data acquired when the student user subsequently takes over the course can be acquired as the result in the sample data to replace the original unqualified result data according to the result statistical rules of the school. For example, the examination score of the lecture a of the student b is 45, and is the type-ineligible score data. Subsequently, when the student B supplementary examination selecting course A, the obtained supplementary examination score is 64. The supplementary examination score 64 is obviously a score obtained after the student b makes an effort to learn the optional lesson a while making full use of the knowledge reserve of the previous basic lesson of the student b with respect to the previous examination score 45. The supplementary test result 64 has smaller accidental errors compared with the previous test result 45, better accords with the true learning potential of the student and has better universality. Therefore, when the sample data is collected, the supplementary examination result 64 obtained when the student B carries out the subsequent supplementary examination and optional repair course A can be used for replacing the previous unqualified result data 45, so that the collected sample data has better universality, and the accuracy of the subsequent model training is further improved.

In order to train and establish a preset lesson selection result prediction model with higher accuracy and smaller error, sample data with the difference degree larger than a preset difference degree threshold (namely sample data with larger contingency, unreasonable property and general rule under the normal condition of a student group cannot be reflected) can be removed from each sample data set in advance by determining and according to the difference degree of the lesson selection results of the sample data set and the preset basic lesson results before the model is trained. Therefore, a sample data set with smaller error and better effect can be obtained. And then, a preset lesson selection result prediction model which is relatively more accurate can be obtained by training the sample data set. The preset difference value degree threshold value can be flexibly set according to specific precision requirements.

In one embodiment, the initial model may further be provided with a dropout parameter. The dropout parameter can be understood as a training parameter in the neural network, and can be used as an adjustment parameter for preventing overfitting when the deep neural network is trained.

In an embodiment, in the process of training the initial model based on the neural network by using the plurality of sample data sets, the method may further include the following steps: and reducing the influence of dirty data on model training by controlling the numerical value of the dropout parameter so as to avoid the over-fitting phenomenon. Therefore, the overfitting problem in model training can be effectively avoided, and the preset selecting and repairing course score prediction model with relatively higher accuracy and relatively better using effect is obtained.

In this embodiment, the dirty data may be specifically understood as data that may introduce a large degree of difference during the model training process, which may cause problems such as model overfitting.

In one embodiment, in order to train to obtain a more detailed preset course selection prediction model corresponding to each specialty, the course score data of the current student user is collected to obtain a large amount of sample data; professional labels of the current students can also be collected. Further, the sample data can be divided into sample data sources corresponding to different professional types according to professional labels of students. Wherein each sample data source corresponds to a specialty. And further, aiming at each specialty, the sample data in the sample data source corresponding to the specialty can be used for establishing a sample data set corresponding to a plurality of selection courses. And aiming at each specialty, respectively training the initial model based on the neural network by using a plurality of sample data sets corresponding to the specialty so as to establish a preset lesson selection result prediction model corresponding to each specialty.

In an embodiment, in the process of training the initial model based on the neural network by using the plurality of sample data sets, the preset repair course performance prediction model meeting the requirements can be obtained by training by adjusting the training times and/or the training time so as to take the training efficiency and the model precision into consideration.

In an embodiment, after a plurality of sample data sets are established according to the sample data, when the method is implemented, the following may be further included: and respectively splitting each sample data set in the plurality of sample data sets into a training subset and a testing subset. When the method is specifically split, 70% of sample data in the sample data set can be randomly extracted to serve as a training subset, and the rest sample data in the sample data set serves as a testing subset.

Further, the initial model may be trained using training subsets of the plurality of sample data sets to obtain a first lesson selection performance prediction model. And then, the first repair course result prediction model can be tested by utilizing a test subset in a plurality of sample data sets to obtain a corresponding test result. And according to the test result, corresponding model parameters in the first course selection result prediction model are adjusted in a targeted manner, and the model is updated, so that an updated model with high accuracy and good use effect can be obtained and used as a preset course selection result prediction model.

By the mode, the accuracy of the established preset repair course result prediction model can be further improved and the model error can be further reduced by performing cross validation on the training subset and the testing subset.

In an embodiment, the repair course recommendation request may further carry a personalized indication parameter of the target user. The personalized indication parameter may specifically include: the indication parameters of the target user for the class time (for example, the requirement of the class time is three weeks, etc.), the indication parameters of the target user for the class place (for example, the requirement of the class place is two teachers, etc.), the indication parameters of the target user for the teacher in class, etc.

Correspondingly, after determining the repair course matched with the target user according to the repair course result prediction result, the server may further obtain associated data (e.g., time information of the lesson, place information of the lesson, teacher information of the lesson, etc.) of the repair course matched with the target user; and further screening the selected and repaired courses matched with the personalized indication parameters of the target user from the selected and repaired courses matched with the target user according to the personalized indication parameters of the target user and the associated data of the selected and repaired courses, and taking the selected and repaired courses matched with the personalized indication parameters of the target user as the target selected and repaired courses to be finally recommended to the target user. Therefore, the repair courses meeting the personalized requirements of the users can be found and recommended for the users, and the use experience of the users is further improved.

In one embodiment, the server may further collect suggestions of current student users during training to establish the preset repair course result prediction model. And then the suggestion of the current student user can be introduced to carry out model training, and a preset repair course score prediction model with better effect is obtained.

Specifically, for example, in the process of training the model by using sample data, the server may adjust the weight of each preset basic course score according to the suggestion of the current student, so that the preset repair course score prediction model can be obtained through more efficient and accurate training.

In an embodiment, after generating and sending recommendation information about the target repair course to the target user, when the method is implemented, the following may be further included: receiving feedback information of a target user aiming at the recommendation information; and modifying and updating the preset repair course result prediction model according to the feedback information.

The feedback information may be forward feedback information, for example, the target user receives the recommendation information and selects the target repair course recommended by the server; or negative feedback information, for example, the target user does not receive the recommendation information, and does not select the target repair course recommended by the server.

In one embodiment, after determining that the feedback information is negative feedback information, the server may specifically modify a relevant model parameter in the preset repair course performance prediction model according to the negative feedback information.

In one embodiment, the server may set a tracking flag on the target user after determining that the feedback information is forward feedback information. And the server can collect and record the real result of the selected course recommended by the target user and the learning server later according to the tracking mark. Subsequently, the real result of the selected course of the target user can be compared with the predicted result of the selected course obtained based on the preset selected course result prediction model to adjust and modify the preset selected course result prediction model in a targeted manner, so that the adjusted and modified preset selected course result prediction model is more accurate.

As can be seen from the above, the method for recommending a repair course provided in the embodiments of the present specification may first retrieve and acquire score data of a basic course preset by a target user according to an identity of the target user carried in a repair course recommendation request initiated by the target user; calling a pre-trained preset election and repair course score prediction model to obtain a corresponding election and repair course score prediction result according to the score data of the basic course preset by the target user; and then, by using the result of predicting the result of the repair course as a basis, determining a repair course with relatively high learning potential of the target user from a plurality of repair courses as a matched repair course, and recommending the repair course to the target user. Therefore, the system can help the student user to efficiently and accurately find the selection and repair course matched with the current learning condition of the student user and suitable for the student user to learn, and the technical problem that the student user cannot efficiently and accurately determine the selection and repair course suitable for the student user is solved.

Embodiments of the present specification further provide a server, including a processor and a memory for storing processor-executable instructions, where the processor, when implemented, may perform the following steps according to the instructions: receiving a course selection and repair recommendation request of a target user; the course selection and repair recommendation request at least carries an identity of a target user; acquiring grade data of a basic course preset by the target user by retrieving a course grade database according to the identity of the target user; calling a preset repair course score prediction model to obtain a corresponding repair course score prediction result according to the score data of the basic course preset by the target user; according to the result of the grading prediction of the repair courses, determining a repair course matched with a target user as a target repair course; and generating and sending recommendation information about the target repair course to the target user.

In order to more accurately complete the above instructions, referring to fig. 2, the present specification further provides another specific server, wherein the server includes a network communication port 201, a processor 202 and a memory 203, and the above structures are connected by an internal cable, so that the structures can perform specific data interaction.

The network communication port 201 may be specifically configured to receive a course selection recommendation request of a target user; and the course selection and repair recommendation request at least carries an identity of a target user.

The processor 202 may be specifically configured to obtain, according to the identity of the target user, result data of a basic course preset by the target user by retrieving a course result database; calling a preset repair course score prediction model to obtain a corresponding repair course score prediction result according to the score data of the basic course preset by the target user; according to the result of the grading prediction of the repair courses, determining a repair course matched with a target user as a target repair course; and generating and sending recommendation information about the target repair course to the target user.

The memory 203 may be specifically configured to store a corresponding instruction program.

In this embodiment, the network communication port 201 may be a virtual port that is bound to different communication protocols, so that different data can be sent or received. For example, the network communication port may be a port responsible for web data communication, a port responsible for FTP data communication, or a port responsible for mail data communication. In addition, the network communication port can also be a communication interface or a communication chip of an entity. For example, it may be a wireless mobile network communication chip, such as GSM, CDMA, etc.; it can also be a Wifi chip; it may also be a bluetooth chip.

In the present embodiment, the processor 202 may be implemented in any suitable manner. For example, the processor may take the form of, for example, a microprocessor or processor and a computer-readable medium that stores computer-readable program code (e.g., software or firmware) executable by the (micro) processor, logic gates, switches, an Application Specific Integrated Circuit (ASIC), a programmable logic controller, an embedded microcontroller, and so forth. The description is not intended to be limiting.

In this embodiment, the memory 203 may include multiple layers, and in a digital system, the memory may be any memory as long as it can store binary data; in an integrated circuit, a circuit without a physical form and with a storage function is also called a memory, such as a RAM, a FIFO and the like; in the system, the storage device in physical form is also called a memory, such as a memory bank, a TF card and the like.

An embodiment of the present specification further provides a computer storage medium based on the recommendation method for a repair course, where the computer storage medium stores computer program instructions, and when the computer program instructions are executed, the computer storage medium implements: receiving a course selection and repair recommendation request of a target user; the course selection and repair recommendation request at least carries an identity of a target user; acquiring grade data of a basic course preset by the target user by retrieving a course grade database according to the identity of the target user; calling a preset repair course score prediction model to obtain a corresponding repair course score prediction result according to the score data of the basic course preset by the target user; according to the result of the grading prediction of the repair courses, determining a repair course matched with a target user as a target repair course; and generating and sending recommendation information about the target repair course to the target user.

In this embodiment, the storage medium includes, but is not limited to, a Random Access Memory (RAM), a Read-Only Memory (ROM), a Cache (Cache), a Hard Disk Drive (HDD), or a Memory Card (Memory Card). The memory may be used to store computer program instructions. The network communication unit may be an interface for performing network connection communication, which is set in accordance with a standard prescribed by a communication protocol.

In this embodiment, the functions and effects specifically realized by the program instructions stored in the computer storage medium can be explained by comparing with other embodiments, and are not described herein again.

Referring to fig. 3, in a software level, an embodiment of the present specification further provides a recommendation device for a repair course. The apparatus may specifically include the following structural modules.

The receiving module 301 may be specifically configured to receive a course selection and repair recommendation request of a target user; the course selection and repair recommendation request at least carries an identity of a target user;

the obtaining module 302 is specifically configured to obtain, according to the identity of the target user, result data of a basic course preset by the target user by retrieving a course result database;

the invoking module 303 may be specifically configured to invoke a preset repair course score predicting model to obtain a corresponding repair course score predicting result according to score data of a basic course preset by the target user;

the determining module 304 may be specifically configured to determine, according to the result of the result prediction of the repair course, a repair course matched with the target user as a target repair course;

the recommending module 305 may be specifically configured to generate and send recommendation information about the target repair course to the target user.

It should be noted that, the units, devices, modules, etc. illustrated in the above embodiments may be implemented by a computer chip or an entity, or implemented by a product with certain functions. For convenience of description, the above devices are described as being divided into various modules by functions, and are described separately. It is to be understood that, in implementing the present specification, functions of each module may be implemented in one or more pieces of software and/or hardware, or a module that implements the same function may be implemented by a combination of a plurality of sub-modules or sub-units, or the like. The above-described embodiments of the apparatus are merely illustrative, and for example, the division of the units is only one logical division, and other divisions may be realized in practice, for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or units, and may be in an electrical, mechanical or other form.

Therefore, the device for recommending the choosing course provided by the embodiment of the specification can help the student user to efficiently and accurately find the choosing course matched with the current learning condition of the student user, is suitable for the choosing course learned by the student user, and solves the technical problem that the student user cannot efficiently and accurately determine the choosing course suitable for the student user.

In one example scenario, the recommendation method for a repair course provided by the embodiment of the present specification may be applied to serve students at university C, and a matching and appropriate repair course is recommended for the students at the school. The following can be referred to for specific implementation.

In this scenario example, the researchers at the college D of C university are selected as service objects, and before the students at the college D of C university enter a stage where they can independently choose courses, they will first repair some basic courses (i.e. preset basic courses) that need to be learned according to the rules of the school, where these courses that need to be repaired include 6 basic courses of higher mathematics i, higher mathematics ii, oil layer physics, seepage mechanics, oil reservoir engineering, oil production engineering, and the like, and the professional courses (i.e. optional courses) that they will independently choose to repair include 6 professional courses of higher oil reservoir engineering, higher seepage mechanics, oil reservoir numerical simulation, enhanced oil recovery method, complex structured well theory and technology, oil production engineering scheme design, and the like. Before implementation, a prediction model for a researcher at college D of university C (i.e., a preset lesson election performance prediction model) may be trained in the following manner.

Step 1: collecting the basic class scores of the current students which must be repaired and the professional class scores which can be selected to be repaired independently.

The method can be used for collecting basic course performance data (including higher mathematics I, higher mathematics II, oil layer physics, seepage mechanics, oil reservoir engineering and oil extraction engineering) which must be repaired and can be independently selected and repaired, and can be used for collecting basic course performance data (including higher oil reservoir engineering, higher seepage mechanics, oil reservoir numerical simulation, a recovery ratio improving method, complex structure well theory and technology and oil extraction engineering scheme design) of previous students of college D of university C.

Step 2: and establishing a prediction model of professional course achievements which can be independently selected for repairing by utilizing a neural network algorithm based on the collected relevant course achievements of the past students.

During the specific training, the following contents can be included:

(1) based on the collected related course achievements of the past students, 6 types of past student achievement data sets (namely a plurality of sample data sets, wherein each sample data set corresponds to one course of the past students) of courses such as high oil deposit engineering, high seepage mechanics, oil deposit numerical simulation, an enhanced oil recovery method, complex structure well theory and technology, oil extraction engineering scheme design and the like are respectively formed according to the standards of selecting and repairing students in the same professional course as one type.

Specifically, for example, a data set of performance of past students in the selection and repair of the high seepage mechanics is taken as an example. As can be seen in fig. 4.

(2) The obtained 6 types of past student performance data are randomly divided into a training set (namely a training subset) and a testing set (namely a testing subset). Wherein, the training set data accounts for 70 percent, and the test set data accounts for 30 percent.

(3) Two parameters, namely a dropout parameter and a training time are set. Wherein the value of the dropout parameter can be set between 0 and 1. In specific implementation, the smaller the value of the dropout parameter is set, the more data points that need to be discarded in the training of the neural network. In addition, a specific numerical value of the training times and a specific numerical value of the training time can be set according to specific conditions.

Experiment verification uses the results of 6 basic courses as independent variables and one alternative course as dependent variables, and a model with higher prediction precision can be obtained by 2500 training times. After 5000 trains the accuracy hardly increases. In the present scenario example, the number of training times is preferably set to 3000.

By adding the dropout parameter, a part of data points in the prediction model can be abandoned, and the over-fitting phenomenon of the prediction model is prevented. This is because when training the model, it is easy to cause the inaccuracy of the prediction model due to the unreasonable quality of the partial data itself. For example: if a student learns well when learning a basic course, but relaxes after selecting a professional course, the independent variable and the dependent variable can not be regularly matched, and if the model takes the data into consideration for training, the overfitting phenomenon of model training can be caused, and an accurate model with universality can not be obtained. Therefore, a dropout parameter is introduced during model training, some bad data can be abandoned in the process of model training, and model accuracy under big data training is improved. The dropout parameter is between 0 and 1, the smaller the data is, the more the data is discarded,

in addition, an interface for the user to freely input is designed in the program code of the training model, so that the user can conveniently input the data according to the data characteristics of the user. Wherein, the interface comprises an interface for freely inputting training times, so that the balance of training precision and training time can be provided for a user. When the model is trained, the more the training times, the higher the training model precision, and in the specific use process, a large number of courses may need to be trained, and the required training time is also longer.

(4) Respectively taking the scores of 6 basic courses of high mathematics I, high mathematics II, oil layer physics, seepage mechanics, oil reservoir engineering, oil extraction engineering and the like which need to be repaired in the 6 types of past student score data as independent variables, and the scores of professional courses which can be independently selected and repaired and correspond to each type of data as a prediction dependent variable, compiling a corresponding program by combining a Python programming language through a neural network algorithm, training a prediction model by utilizing training set data, and testing the model by utilizing testing set data. Through the training and the testing, the testing error of the model is globally minimized, so that a prediction model capable of predicting professional performances of high reservoir engineering, high seepage mechanics, reservoir numerical simulation, an enhanced recovery method, complex structure well theory and technology, oil extraction engineering scheme design and the like is obtained respectively, and the adopted calculation program code can be referred to as shown in figure 5.

After training to obtain the prediction model, the prediction model can be used to help the researchers at college D of university C select the appropriate course for repair. Specifically, the following steps may be included.

And step 3: acquiring basic course achievements (namely achievement data of basic courses preset by target users) which are required to be repaired and exist by current course selection students (namely target users).

And 4, step 4: and based on the established prediction model of the professional class achievements which can be selected and reviewed independently, predicting the professional class achievements which can be selected and reviewed independently corresponding to the current students according to the existing basic class achievements which must be reviewed by the current students.

The method specifically comprises the following steps: the method comprises the steps of taking the basic course achievements which are required to be repaired of students in the current course selection as input feature vectors of a model, and respectively obtaining predicted values (namely the result of predicting the course selection achievements) of all the professional courses which can be independently selected to be repaired and learned, corresponding to the students in the current course, according to established prediction models of the professional course achievements of high oil reservoir engineering, high seepage mechanics, oil reservoir numerical simulation, a recovery ratio improving method, complex structure well theory and technology, oil extraction engineering scheme design and the like. The predicted results can be seen in FIG. 6.

And 5: and (4) according to the predicted current student performance, the students can independently select professional lesson achievements for repairing and recommending lessons.

The method specifically comprises the following steps: according to the predicted professional class scores of all current students capable of selecting and repairing independently, firstly, for each class independently, the predicted score can enable the students to learn the class by themselves and finally obtain the scores, the higher the score is, the more suitable the students can learn the class, the more the students can learn the ability of the class and the potential of learning the class, taking a schoolmate as an example, the 6 professional class scores of the schoolmate which can be selected and repaired are all more than 80, the more firm basic knowledge of the schmate is described, the stronger comprehensive ability is realized, and all the classes can obtain good scores; then, for all professional classes that can be selected to be reviewed autonomously, the scores are ranked in the order from big to small, taking a certain classmate as an example, the rank of the predicted scores of the optional review classes can be shown in fig. 7, the oil extraction engineering project design discipline is ranked first, the predicted score is the largest, which shows that in the 6 professional classes that can be selected to be reviewed, the student should preferentially select the class to learn, and the largest score can be obtained.

In the scene example, a course selection recommendation method based on big data analysis is established by utilizing a neural network algorithm based on course selection data of a large number of students generated by long-term accumulation, scientific course selection recommendation can be effectively provided for college students, the specialties and the potential abilities of the students are fully played, the students are helped to learn useful knowledge, the students are not puzzled on how to select courses, the students are not blindly helped to study in order to fill up the school score, the vigor is fully put on the study, and a foundation is laid for the improvement of a course selection system of college schools.

Although the present specification provides method steps as described in the examples or flowcharts, additional or fewer steps may be included based on conventional or non-inventive means. The order of steps recited in the embodiments is merely one manner of performing the steps in a multitude of orders and does not represent the only order of execution. When an apparatus or client product in practice executes, it may execute sequentially or in parallel (e.g., in a parallel processor or multithreaded processing environment, or even in a distributed data processing environment) according to the embodiments or methods shown in the figures. The terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, the presence of additional identical or equivalent elements in a process, method, article, or apparatus that comprises the recited elements is not excluded. The terms first, second, etc. are used to denote names, but not any particular order.

Those skilled in the art will also appreciate that, in addition to implementing the controller as pure computer readable program code, the same functionality can be implemented by logically programming method steps such that the controller is in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers, embedded microcontrollers and the like. Such a controller may therefore be considered as a hardware component, and the means included therein for performing the various functions may also be considered as a structure within the hardware component. Or even means for performing the functions may be regarded as being both a software module for performing the method and a structure within a hardware component.

This description may be described in the general context of computer-executable instructions, such as program modules, being executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, classes, etc. that perform particular tasks or implement particular abstract data types. The specification may also be practiced in distributed computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules may be located in both local and remote computer storage media including memory storage devices.

From the above description of the embodiments, it is clear to those skilled in the art that the present specification can be implemented by software plus necessary general hardware platform. With this understanding, the technical solutions in the present specification may be essentially embodied in the form of a software product, which may be stored in a storage medium, such as a ROM/RAM, a magnetic disk, an optical disk, etc., and includes several instructions for enabling a computer device (which may be a personal computer, a mobile terminal, a server, or a network device, etc.) to execute the methods described in the embodiments or some parts of the embodiments in the present specification.

The embodiments in the present specification are described in a progressive manner, and the same or similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments. The description is operational with numerous general purpose or special purpose computing system environments or configurations. For example: personal computers, server computers, hand-held or portable devices, tablet-type devices, multiprocessor systems, microprocessor-based systems, set top boxes, programmable electronic devices, network PCs, minicomputers, mainframe computers, distributed computing environments that include any of the above systems or devices, and the like.

While the specification has been described with examples, those skilled in the art will appreciate that there are numerous variations and permutations of the specification that do not depart from the spirit of the specification, and it is intended that the appended claims include such variations and modifications that do not depart from the spirit of the specification.

Claims

1. A recommendation method for a repair course is characterized by comprising the following steps:

2. The method of claim 1, wherein the repair course recommendation request further carries a professional label of the target user.

3. The method of claim 2, further comprising:

4. The method of claim 1, wherein the pre-defined repair session performance prediction model is built as follows:

5. The method of claim 4, wherein prior to training the initial neural network-based model with the plurality of sample data sets, respectively, the method further comprises:

6. The method of claim 4, wherein the initial model comprises a plurality of initial network model structures, and wherein each of the plurality of initial network model structures corresponds to a repair course.

7. The method of claim 4, wherein in training the initial neural network-based model using the plurality of sample data sets, respectively, the method further comprises:

8. The method of claim 1, wherein determining a repair course matching a target user based on the repair course achievement prediction comprises:

9. A recommendation device for a repair course, comprising:

10. A server comprising a processor and a memory for storing processor-executable instructions which, when executed by the processor, implement the steps of the method of any one of claims 1 to 8.