CN112417002B - Information literacy data mining method and system applied to education informatization - Google Patents

Abstract

The invention discloses an information literacy data mining method, system, computer equipment and storage medium applied to education informatization. The method includes the steps of: collecting the information literacy test data, index structure data and index parameter data input by the testee; constructing an information literacy model generation program, and the information literacy model generation program is used to receive the index structure data and the index Parameter data, generate an information literacy model based on the Bayesian network according to the index structure data and the index parameter data; construct an information literacy data mining program, and the information literacy data mining program is used to receive the information literacy model and test Data, output information literacy test results. The invention can accurately express the dependency relationship between complex data with computer language, has strong interpretation ability, thereby improving the accuracy of data mining and high data value.

Description

Information literacy data mining method and system applied to educational informatization

Technical Field

The present invention belongs to the technical field of data mining, and more specifically, relates to an information literacy data mining method and system applied to educational informatization.

Background Art

Under the wave of rapid development of informatization, the field of education has been fully integrated with computer technology and evolved into a variety of forms. Different from the previous data analysis that relied on manual experience, the modern information literacy data mining method uses computer technology to fully mine valuable information from massive test big data, realizes the visualization and automatic analysis of measurement data, and is a specific application of data mining technology in the field of educational informatization. It has a wide range of application scenarios and extremely high application value. However, in the existing information literacy data mining methods, the data mining model is simple and lacks flexibility. It is impossible to accurately express the dependency relationship between complex data in computer language, and the interpretation ability is not strong, which leads to the fact that the mined data cannot accurately and objectively reflect the real situation of the data, and the data value is low.

Summary of the invention

In response to at least one defect or improvement need in the prior art, the present invention provides an information literacy data mining method, system, computer equipment and storage medium applied to educational informatization, which can accurately express the dependencies between complex data in computer language and has strong interpretation ability, thereby improving the accuracy of the mined data and increasing the data value.

To achieve the above-mentioned purpose, according to a first aspect of the present invention, there is provided an information literacy data mining method applied to educational informatization, comprising:

Collecting information literacy test data input by the testee, and storing the test data as a score data matrix table;

Collect the indicator structure data input by the expert, and store the indicator structure data in JSON format;

Collect the indicator parameter data input by the expert, and store the indicator parameter data in JSON format;

Constructing an information literacy model generation program, wherein the information literacy model generation program is used to receive the indicator structure data and the indicator parameter data, and generate an information literacy model based on a Bayesian network according to the indicator structure data and the indicator parameter data;

An information literacy data mining program is constructed, wherein the information literacy data mining program is used to receive the information literacy model and the score data matrix table, and output information literacy test results.

Preferably, the information literacy test data includes a plurality of test items, and each test item includes a plurality of levels;

The performance data matrix table is an n×m matrix, where n is the total number of test subjects and m is the total number of test items. Each row of the performance data matrix table represents the test data of all test items of a test subject, and each column of the performance data matrix table represents the test data of all test subjects of a test item.

Preferably, the indicator structure data includes test items, indicators represented by each test item, and indicator dependencies, and storing the indicator structure data in JSON format includes the steps of:

The test item or indicator is used as the key of the JSON object, and the indicator represented by the test item or the indicator on which the indicator depends is used as the value of the JSON object.

Preferably, the indicator parameters include a priori probability parameters and conditional probability parameters, the priori probability parameters refer to the probability values of indicators without dependencies at each level, and the conditional probability parameters refer to the probability values of indicators with dependencies at each level, and storing the indicator parameter data in JSON format includes the steps of:

The indicators without dependency are stored as keys of a priori probability parameter object, and the probability values of the indicators without dependency at each level are stored as values of the priori probability parameter object;

The indicators or test items with dependencies are stored as the keys of the conditional probability parameter objects, the values of which are the lists of dependent objects, the keys of the dependent objects are the dependent indicators, and the values of the dependent objects are the probability values of the dependent indicators at each level.

Preferably, the generating of the information literacy model based on the Bayesian network comprises the steps of:

According to the indicator structure data and the dependent parameter data, a tail-to-tail connection method is adopted to construct an information literacy model based on a Bayesian network, and a linked list is used to represent the Bayesian network, wherein the nodes of the Bayesian network represent test items or indicators, and multidimensional data is used in the nodes to store the indicator parameter data corresponding to the node indicator or test item.

Preferably, using a linked list to represent the Bayesian network includes the following steps:

A linked list is used to construct a node array, each node array includes three fields, the first field stores the name of the test item or indicator represented by the node, the second field stores a pointer to a multidimensional array of indicator parameter data corresponding to the node, and the third field stores a pointer to the object on which the node depends. Preferably, the output of the information literacy test result includes the steps of:

Input the Bayesian information literacy model and the test data, calculate the probability of each level of the test-taker's information literacy according to the Bayesian formula, and finally select the level with the highest probability as the final result of the test-taker.

According to a second aspect of the present invention, there is provided an information literacy data mining system for educational informatization, comprising:

A test data collection module, used to collect information literacy test data input by the testee, and store the test data as a score data matrix table;

An indicator structure data collection module, used to collect indicator structure data input by experts and store the indicator structure data in JSON format;

An indicator parameter data collection module, used to collect indicator parameter data input by experts and store the indicator parameter data in JSON format;

A model building module, used to build an information literacy model generation program, wherein the information literacy model generation program is used to receive the indicator structure data and the indicator parameter data, and generate an information literacy model based on a Bayesian network according to the indicator structure data and the indicator parameter data;

The data mining module is used to construct an information literacy data mining program, wherein the information literacy data mining program is used to receive the information literacy model and the score data matrix table and output the information literacy test result.

According to a third aspect of the present invention, there is provided a computer device comprising a processor and a memory, wherein the processor runs a program corresponding to the executable program code by reading the executable program code stored in the memory, so as to implement any one of the above methods.

According to a fourth aspect of the present invention, there is provided a computer-readable storage medium having a computer program stored thereon, wherein the computer program implements any one of the above methods when executed by a processor.

In general, compared with the prior art, the present invention has beneficial effects: an information literacy model is generated based on indicator structure data and indicator parameter data, and the Bayesian network is used in the information literacy model to represent the indicator structure and indicator parameters, which can accurately express the dependency relationship between complex data in computer language and has strong interpretation ability, thereby improving the accuracy of data mining and increasing the data value.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 is a schematic diagram of the principle of the information literacy data mining method according to an embodiment of the present invention;

FIG2 is a schematic diagram of an information literacy model according to an embodiment of the present invention;

FIG3 is a schematic diagram of a node array according to an embodiment of the present invention;

FIG. 4 is a schematic diagram of a link table according to an embodiment of the present invention.

DETAILED DESCRIPTION

In order to make the purpose, technical scheme and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

The information literacy data mining method applied to educational informatization in the embodiment of the present invention is based on the use of computer technology to fully mine valuable information from massive test big data, realize visualization and automatic analysis of measurement data, and is a specific application of data mining technology in the field of educational informatization. The above-mentioned information literacy data mining method can be implemented on a server or a computer terminal.

An information literacy data mining method applied to educational informatization according to an embodiment of the present invention includes steps S1 to S5.

S1, collects the information literacy test data input by the test takers, and stores the test data as a score data matrix table.

Preferably, the information literacy test data includes multiple test items, each test item includes multiple levels, for example, the levels range from A to Z. The test data of multiple test subjects are summarized, and the score data matrix table is an n×m matrix, n is the total number of test subjects, m is the total number of test items, each row of the score data matrix table represents the test data of all test items of one test subject, and each column of the score data matrix table represents the test data of all test subjects of one test item.

For example, test item 1 is a multiple-choice question 1, which is a question about EXCEL, reflecting the subject's information skills and information application abilities. This test item includes 2 levels, A for a correct answer, and B for a wrong answer. Test item 1 is a scale question 2, with 4 options, representing 4 levels A, B, C, and D, indicating four descending levels. It asks the subject about the use of mobile phones, reflecting the subject's information habits.

The following table is an example of a performance data matrix.

Test subjects Test Item 1 Test Item 2 1 A B 2 B B

S2, collects the indicator structure data input by experts and stores the indicator structure data in JSON format.

Indicator structure data refers to the data that defines the test items and indicator structure system.

Preferably, the indicator structure data includes test items, indicators represented by each test item, and indicator dependencies. Storing the indicator structure data in JSON format includes the steps of using the test item or indicator as the key of the JSON object (dependencies), and using the indicator represented by the test item or the indicator on which the indicator depends as the value of the JSON object.

The following is a specific example:

For example, if single-choice question 1 reflects the subject's information skills, it will be recorded as:

"Single-choice question 1": ["Information skills", "Information perception"]

The information skills index is a secondary index of information knowledge and skills, which can be expressed as:

“Information skills”: [“Information knowledge and skills”]

Preferably, the indicator structure data input by multiple experts can be collected. When the indicator structure judgment data of multiple experts are inconsistent, a consensus method can be pre-defined to obtain unique indicator structure data according to the pre-defined consensus method. For example, if half or more than half of the experts give support for the dependency relationship between "test item 2" and "indicator 3", it can be retained, otherwise it will not be retained.

S3 collects the indicator parameter data input by experts and stores the indicator parameter data in JSON format.

The indicator parameter data is the definition data of the probability value of each level of the indicator or test item.

Preferably, the indicator parameters include prior probability parameters and conditional probability parameters. The prior probability parameters refer to the probability values of indicators at each level without dependencies, and the conditional probability parameters refer to the probability values of indicators at each level with dependencies. Storing the indicator parameter data in JSON format includes steps.

The indicators without dependencies are stored as the keys of the prior probability parameter object (prior_probabilities), and the probability values of each level of the indicators without dependencies are stored as the values of the prior probability parameter object;

The indicators or test items with dependencies are stored as the keys of the conditional probability parameter object (conditional_probabilities). The value of the conditional probability parameter object is a list of dependent objects. The key of the dependent object is the dependent indicator, and the value of the dependent object is the probability value of the dependent indicator at each level.

The following is a specific example of a priori probability parameters, where A, B, C, and D represent different levels of indicators:

When the opinions of multiple experts are inconsistent, a consensus method can be pre-defined to obtain unique indicator structure data according to the pre-defined consensus method. For example, taking the average value, assuming that different experts judge the probability values of indicator 4 as level A as 0.15, 0.25, 0.2, 0.1, and 0.5, the average value is 0.15.

The following is a specific example of a conditional probability parameter.

For example, when the expert judges that the subject's information skill level is A and the information application level is B, the probability of single-choice question 1 being correct is 0.8, which can be expressed as:

S4, constructing an information literacy model generation program, the information literacy model generation program is used to receive indicator structure data and indicator parameter data, and generate an information literacy model based on a Bayesian network according to the indicator structure data and indicator parameter data.

According to the indicator structure data and dependent parameter data, each test item and indicator actually constitutes a one-way dependent Bayesian network. Each node of the network represents a test item or indicator, which is connected in a tail-to-tail manner, and the arrows represent their dependencies, as shown in Figure 2. Figure 2 is a specific example of an information literacy model based on a Bayesian network.

The Bayesian-based information literacy model generation program inputs the indicator structure data and dependency parameter data. First, the Bayesian network structure diagram is represented by a computer data structure according to the structure. The structure diagram is stored in an adjacency list (linked) representation, as shown in Figure 3. Extract the information in the indicator structure data and use a linked list to build a node array. The node array consists of the following node structure. The structure contains three fields. The first field stores the name of the test item or indicator represented by the node. The second field stores a pointer to the multidimensional array of the indicator parameter data corresponding to the node, and the third field stores a dependency pointer to the object on which the node depends. The dependency pointer points to a dependency structure. The first field of the dependency structure records the position of the dependent node in the array, that is, the offset relative to the array head, and the second field of the dependency structure records the remaining dependency structures. Extract the key-value pairs in the indicator structure data one by one. If the key is not in the node of the node array, insert a node structure after the node array. Then read its dependency. If the dependency is already in the array, add a dependency structure directly after the key unit structure; if the dependency is not in the array, first add a node structure to the array, and then add a dependency structure.

FIG. 4 is a schematic diagram showing the Bayesian network of FIG. 2 represented by an adjacency list.

Then extract the indicator parameter data and construct a multidimensional array for each node. The processing method of the prior parameters and dependent parameters is the same. The key corresponds to the node name, the array value is stored in the list value, the array dimension includes the dependent dimension and the value dimension, and the number of dimensions is the number of dependencies plus 1. The length of the dependent dimension is the number of levels it depends on, and the length of the value dimension is the number of node value levels. Convert the levels of A, B, C... to coordinates of 0, 1, 2... and use the probability value as the value of the corresponding position in the multidimensional array. If you put the following object into

的相应表示为M[0,1,0]＝0.8。最后将M的指针放入相应节点结构的数组指针字段中。

The corresponding representation is M[0,1,0] = 0.8. Finally, the pointer of M is placed in the array pointer field of the corresponding node structure.

The output adjacency list is used as a Bayesian-based information literacy data mining model.

S5, construct an information literacy data mining program, which is used to receive the information literacy model and the performance data matrix table and output the information literacy label.

The data mining program based on the Bayesian information literacy model takes as input the Bayesian information literacy model and the score table, and calculates the information literacy probabilities of the test takers according to the Bayesian formula:

First, expand the formula based on the Bayesian network. For example, based on the Bayesian network in Figure 2, e and d are test item 1 and test item 2, respectively, and a, c, and b are indicators 1, indicator 2, and indicator 3. If we know that the test subject is grade A in test item 1 and grade C in test item 2, and we want to find the probability that the test subject is grade A in indicator 1, that is, the probability of a = A when d = A and e = C, the formula is as follows:

b and c can take any level, so they are set to X and Y. The formula is reduced to the multiplication of the probability of the node at each level of its dependency.

Secondly, query the value in the Bayesian network according to the formula. As in the above example, X, Y represent any level of b, c, P(a＝A) is the prior probability, which is stored in the index 1 node, and P(b|a) is stored in the index 2 node. Similarly, P(e|a, b) is stored in the test item 1 node. For example, the value of P(e＝C|a＝A, c＝B) is at the [0,1,2] position of the multidimensional array _Me of the test item 1 node.

Extract the probability values in the relevant nodes to get the value of a situation.

P(a＝A,b＝A,c＝A,d＝A,e＝C)＝

P(a＝A)P(c＝A|a＝A)P(b＝A|a＝A)P(e＝C|a＝A,b＝A)P(d＝A|b＝A, c＝A)

Then add up the values for each case.

That is, we get the probability of the index 1 node level A. We calculate the probabilities of different levels of the index 1 node and select the level with the highest probability as the final result of the tester.

Finally, extract the subject number column of the input score data matrix table, add a column to record the model judgment result of each subject, and output the modified score data matrix table as the final output, as follows.

Test subjects score 1 A 2 B

An information literacy data mining system for educational informatization according to an embodiment of the present invention is characterized by comprising:

A test data collection module is used to collect information literacy test data input by the testee and store the test data as a score data matrix table;

The indicator structure data collection module is used to collect the indicator structure data input by experts and store the indicator structure data in JSON format;

The indicator parameter data collection module is used to collect the indicator parameter data input by experts and store the indicator parameter data in JSON format;

A model building module is used to build an information literacy model generation program, and the information literacy model generation program is used to receive indicator structure data and indicator parameter data, and generate an information literacy model based on a Bayesian network according to the indicator structure data and the indicator parameter data;

The data mining module is used to construct an information literacy data mining program. The information literacy data mining program is used to receive an information literacy model and a performance data matrix table and output an information literacy label.

The implementation principle and technical effects of the system are similar to the above method and will not be repeated here.

An embodiment of the present invention also provides a computer device, including a processor and a memory, wherein the processor runs a program corresponding to the executable program code by reading the executable program code stored in the memory, so as to implement any of the above-mentioned information literacy data mining methods.

The embodiment of the present invention further provides a computer-readable storage medium on which a computer program is stored, and the computer program is executed by a processor to implement the technical solution of any of the above-mentioned information literacy data mining method embodiments. Its implementation principle and technical effect are similar to those of the above-mentioned method, and will not be repeated here.

It must be noted that in any of the above embodiments, the method is not necessarily executed in sequence according to the sequence number. As long as it cannot be inferred from the execution logic that it must be executed in a certain order, it means that it can be executed in any other possible order.

It will be easily understood by those skilled in the art that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims (7)

1. An information literacy data mining method applied to education informatization is characterized by comprising the following steps:

collecting information literacy test data input by a tested person, and storing the test data as a result data matrix table;

acquiring index structure data input by an expert, and storing the index structure data into a JSON format;

acquiring index parameter data input by an expert, and storing the index parameter data into a JSON format;

constructing an information literacy model generation program, wherein the information literacy model generation program is used for receiving the index structure data and the index parameter data and generating an information literacy model based on a Bayesian network according to the index structure data and the index parameter data;

constructing an information literacy data mining program, wherein the information literacy data mining program is used for receiving the information literacy model and the achievement data matrix table and outputting an information literacy test result;

the index structure data refers to data defining a test item and an index structure system; the index parameter data is definition data of probability values of indexes or test items in all levels;

the method for generating the Bayesian network-based information literacy model comprises the following steps:

according to the index structure data and the index parameter data, an information literacy model based on a Bayesian network is constructed in a tail-to-tail connection mode, and the Bayesian network is represented by a chain table, wherein nodes of the Bayesian network represent test items or indexes, and multidimensional data are used in the nodes to store the index of the node or the index parameter data corresponding to the test items;

the method for representing the Bayesian network by using the chain table comprises the following steps:

the method comprises the steps that a chained list is adopted to construct node arrays, each node array comprises three fields, the first field stores names of test items or indexes represented by nodes, the second field stores pointers of multidimensional arrays pointing to index parameter data corresponding to the nodes, and the third field stores pointers pointing to objects depended by the nodes;

the index structure data comprises test items, indexes represented by each test item and index dependency relations, and the step of storing the index structure data into a JSON format comprises the following steps:

and taking the test item or the index as a key of the JSON object, and taking the index represented by the test item or the index depended by the index as the value of the JSON object.

2. The method of information literacy data mining applied to educational informatization of claim 1, wherein the information literacy test data comprises a plurality of test items, each test item comprising a plurality of levels;

the result data matrix table is an n multiplied by m matrix, n is the total number of testees, m is the total number of test items, each row of the result data matrix table represents the test data of all the test items of one tester, and each column of the result data matrix table represents the test data of all the testees of one test item.

3. The method as claimed in claim 2, wherein the index parameters include a prior probability parameter and a conditional probability parameter, the prior probability parameter is a probability value of each level of an index without dependency, the conditional probability parameter is a probability value of each level of an index with dependency, and the step of storing the index parameter data in JSON format comprises the steps of:

storing the index without the dependency relationship as a key of a prior probability parameter object, and storing the probability value of the index without the dependency relationship as each grade as the value of the prior probability parameter object;

and storing the indexes or test items with the dependency relationship as keys of a conditional probability parameter object, wherein the value of the conditional probability parameter object is a list of the dependent objects, the keys of the dependent objects are the dependent indexes, and the values of the dependent objects are the probability values of all levels of the dependent indexes.

4. The method for information literacy data mining applied to educational informatization according to claim 3, wherein the output of the information literacy test result comprises the steps of:

and inputting an information literacy model and test data based on Bayes, calculating the probability of each grade of the information literacy of the tester according to a Bayes formula, and finally selecting the grade with the highest probability as the final result of the tester.

5. An information literacy data mining system applied to educational informatization, comprising:

the test data acquisition module is used for acquiring information literacy test data input by a tested person and storing the test data into a score data matrix table;

the index structure data acquisition module is used for acquiring index structure data input by experts and storing the index structure data into a JSON format;

the index parameter data acquisition module is used for acquiring index parameter data input by experts and storing the index parameter data into a JSON format;

the model building module is used for building an information literacy model generation program, the information literacy model generation program is used for receiving the index structure data and the index parameter data and generating an information literacy model based on a Bayesian network according to the index structure data and the index parameter data;

the data mining module is used for constructing an information literacy data mining program, and the information literacy data mining program is used for receiving the information literacy model and the score data matrix table and outputting an information literacy test result;

the index structure data refers to data defining a test item and an index structure system; the index parameter data is definition data of probability values of indexes or test items in all levels;

the method for generating the Bayesian network-based information literacy model comprises the following steps:

according to the index structure data and the index parameter data, an information literacy model based on a Bayesian network is constructed in a tail-to-tail connection mode, and the Bayesian network is represented by a chain table, wherein nodes of the Bayesian network represent test items or indexes, and multidimensional data are used in the nodes to store the index of the node or the index parameter data corresponding to the test items;

the method for representing the Bayesian network by using the chain table comprises the following steps:

the method comprises the steps that a chained list is adopted to construct node arrays, each node array comprises three fields, the first field stores names of test items or indexes represented by nodes, the second field stores pointers of multidimensional arrays pointing to index parameter data corresponding to the nodes, and the third field stores pointers pointing to objects depended by the nodes;

the index structure data comprises test items, indexes represented by each test item and index dependency relations, and the step of storing the index structure data into a JSON format comprises the following steps:

and taking the test item or the index as a key of the JSON object, and taking the index represented by the test item or the index depended by the index as the value of the JSON object.

6. A computer device comprising a processor and a memory, wherein the processor runs a program corresponding to an executable program code stored in the memory by reading the executable program code for implementing the method according to any one of claims 1 to 4.

7. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the method according to any one of claims 1 to 4.

Images