KR20020091335A - data sorting method of problem DB to change a degree of difficulty - Google Patents

Abstract

PURPOSE: A method for arranging data of a question database is provided to delete questions which degrade a discrimination of a learner, classify questions in accordance with a real difficulty degree of learners according to grades, and readjust the questions. CONSTITUTION: A learner performs a learning through a terminal, a control server, and various kinds of databases on the Internet, and learned result is stored in the database. A learning passing time is counted and checks whether the time is passed over one month(S1). If the time is passed over one month, one question is selected from a representative question database and an adaptation question database(S2) and a plurality of correct answer rate distributions is searched from a solve result database. If a correct answer rate is less than 30%(S3), the grade is allocated as the superior. If a correct answer rate is more than 70%(S5), the grade is allocated as the inferior(S7). A difficulty degree field is modified in the corresponding question databases and readjusted, and changes a stored memory address, if necessary(S8). If a correct answer rate of the upper group 30% out of the total learners of a readjusted question is less than 60%(S9) or a correct answer rate of the lower group 30% out of the total learners of the readjusted question is more than 40%(S10), the question is withdrawn(S11). If a question to be arranged exists(S12), the above stages(S2-S11) are repeated for readjusting a difficulty degree thereof.

Description

Data sorting method of problem DB to change a degree of difficulty}

The present invention relates to a data cleansing method of a problem database for difficulty shifting, and more particularly, to a data cleansing method of problem databases for difficulty shifting which is rearranged for each difficulty level of the problem data stored in the problem database.

In general, the problem database stores a large number of questions for learning by type and provides them to the learner via the Internet when necessary.

However, such a problem database is not determined the difficulty of the problem, and because it is stored in a random order, a number of problems with a low discrimination ability is permanently stored, there is a problem that can not improve the education according to learning.

In addition, if the problem of low discrimination power is stored more and more, there is a problem in that the actual storage area of the database becomes relatively small.

The present invention has been made to solve the above problems, the difficulty of removing the problem of dropping the discrimination of learners among the problems stored in the problem database and the difficulty of classifying the problem by classifying the problem according to the learners real difficulty To provide a method for data cleansing of problem databases, the purpose is to.

The object of the present invention is to count the elapsed time learned by the at least one learner connected to the Internet and compare it to a reference elapsed time, and if the reference elapsed time is exceeded, at least one stored in a problem database. Sequentially selecting and re-adjusting the difficulty level by setting the difficulty level to upper, middle, and lower by retrieving a plurality of correct answer rate distributions from the problem-solving result database according to the problem; From the information database, if the upper group correct answer rate of a certain area among the total learners is relatively low, and the lower group correct answer rate of a certain group among the total learners is high, then it is determined that there is no discrimination in the learning, and the problem is eliminated. Difficulty comprising the step of sequentially repeating the readjustment step and exit step It is achieved by the data cleanup method of problem database for moving the diagram.

Here, the reference elapsed time is preferably based on any one of a week, full, monthly, quarterly. Preferably, the upper group and the lower group belong to at least 30% or more from the upper and lower parts of the whole.

1 is a system configuration for implementing a data cleaning method of a problem database according to the present invention,

2 is a flowchart illustrating a data cleaning method of a problem database for difficulty movement according to the present invention.

Explanation of symbols on main parts of drawing

10; Terminal 20 Control Server

30; Representative Problem Database 40 Adaptive Problem Database

50; Problem solving results database 60; Member Information Database

Hereinafter, with reference to the accompanying drawings will be described in detail the data cleaning method of the problem database for the difficulty movement according to the present invention. 1 is a system configuration for implementing a data cleaning method of a problem database according to the present invention, Figure 2 is a flowchart showing a data cleaning method of a problem database for the difficulty movement according to the present invention.

As shown in FIG. 1, the system configuration for the data cleaning method of the problem database is provided with a user interface for establishing a client web browser to access the Internet, and accessing the learning site desired by the learner through the interface. Learner terminal 10 for performing the learning, and connected to the terminal 10 via the Internet network and controls a plurality of databases for one-to-one learning to achieve the desired learning by the learner, the databases It has a control server 20 that performs simultaneously so that the functions of organic data exchange, update, new establishment, and deletion do not interfere with each other.

The control server 20 stores probable problems distinguished by the difficulty level of the problem and the representative types of the problem according to the identification code and the code. The control server by dividing the representative problem database 30 to be rearranged to at least one of the deletion of the loss problem, and the modified variation problems applicable to the learner from at least one of the representative problems for further study by difficulty level The adaptation problem database 40, which is stored retrievable by the instruction of (20) and rearranged to at least one of the difficulty updating problem and the elimination of discrimination problem among the above problems, is connected. Here, the databases (30, 40) is free to move the data between each other, it is configured to perform the role of the task database operated by the teacher while serving as a simple problem database.

In addition, the control server 20 temporarily stores the learning answer of the learner solved the problem received by the learner through the terminal 10, and compares the solution of the problem with the answer to the learner answer to store the solution result Member information database that stores the results database 50, the learner's personal information and personal information and learning information, that is, learning achievement information, learning progress information, teacher assignment information by the learner's ID or name and specific number ( 60) are connected in random order. Here, the databases 50 and 60 are free to move data between each other.

In particular, the control server 20 retrieves the learner's learning achievement information from the member information database 60 and calculates a specific upper group by recognizing the ranking according to the total learner's grade therefrom, and the correct answer rate of the upper group. The first problem which is 60% or less is searched from the problem-specific solving database 50. At the same time, a specific subgroup is calculated and when the second problem with a correct answer rate of more than 40% of the fraud subgroup is the same as the first problem, it is withdrawn from the problem databases 30 and 40. Here, the 60% and the 40% is the correct answer rate by more than 30% learners of the lower group and the upper group, respectively.

And if the correct answer rate is relatively small from the problem-specific solve database 50, the difficulty level is set to upper, and if the correct answer rate is relatively middle, the difficulty level is set to medium, and if the correct answer rate is relatively high, the difficulty level is set to be lower. The difficulty adjustment and problem withdrawal are made based on the results of learning by unspecified learners made for a certain time, that is, weekly, full, monthly, quarterly.

As described above, in the data cleaning method of the problem database, as illustrated in FIG. 2, a learner first learns through the Internet through the terminal 10, the control server 20, and various databases (not shown). The results of the training are stored in the database. The elapsed time learned in this way is counted to compare whether the reference elapsed time, that is, one week, full, one month, or one month of the quarter is reached (S1). Here, if the elapsed learning time does not exceed the reference elapsed time (hereinafter, referred to as one month), the comparison is continued until it exceeds.

When the learning elapsed time is exceeded in one month, any one of the problems learned so far is selected from the representative problem database 30 and the adaptive problem database 40 (S2), according to the problem-specific solution results Retrieve a plurality of correct percentage distributions from the database 50. For example, if the correct answer rate is less than 30% (S3) is determined that the difficulty of the corresponding problem belonging to any one of the corresponding problem database (30,40) has a very high discriminating power and assigns its grade to the award ( S4), when the correct answer rate is less than 30% to 70% (S5) determines that the difficulty of the problem belonging to any one of the corresponding problem database (30,40) is somewhat discriminating and assigns his grade to medium (S6). When the correct answer rate is 70% or more (S5), it is determined that there is a relatively low discrimination power of a problem belonging to any one of the problem databases 30 and 40, and the grade thereof is assigned below (S7).

As described above, the difficulty fields may be modified and adjusted in the problem databases 30 and 40 corresponding to the problem difficulty, and the stored memory address may be changed if necessary (S8). The difficulty level of each difficulty in which the difficulty field has been readjusted is 30% of the correct answer rate of the upper group 30% corresponding to a certain area among the total learners from the member information database 60 or less (S9), or a certain area among the total learners. If the correct answer rate of 30% of the sub-group corresponding to more than 40% (S10) determines that there is no discrimination about learning and eliminate the problem (S11). That is, the problem that 30% of the upper group fails to match and 30% of the lower group does not fit well is judged as a problem in which the problem itself is not established and the discriminating power is lost.

Then, if the control server 20 has more problems to search and clean up the problem databases (30, 40) (S12), the difficulty by repeating the second step to the eleventh step until there is no problem to clean up Reschedule and delete the issue that corresponds to the exit.

As a result, the learner learns with a new difficulty level every time the reference elapses, that is, one month, thereby improving the learning ability.

The present invention deletes the problems that reduce the discrimination of learners among the problems stored in the problem database and classifies the problems according to the learners' real difficulty by classifying and realigning the learners according to the real difficulty so that the learners receive the correct learning evaluation according to the real difficulty. It can provide the ability to learners correctly and has the effect of maximizing the storage efficiency of problem databases.

Claims (3)

Counting elapsed time learned by at least one learner connected to the Internet and comparing the elapsed time to a reference elapsed time; When the reference elapsed time is exceeded, at least one problem stored in the problem database is sequentially selected, and according to the result-based problem solving database for each problem, multiple answer rate distributions are searched and the difficulty level is set to high, medium, or low and readjusted. Making a step; At least one of the re-leveled questions is determined by the member information database to have a relatively low upper group correct answer rate in a certain area among the total learners, and a lower correct percentage rate for a lower group in the certain area among the total learners is determined to have no discrimination in learning. To retire the problem; And a next step of repeating the readjustment step and the retirement step sequentially. The method of claim 1; The reference elapsed time is a data cleaning method of a problem database for difficulty movements, characterized in that based on any one of a week, full, monthly, quarterly period. The method of claim 1; And the upper group and the lower group belong to at least 30% or more from the upper part and the lower part of the whole.