Feasibility study of establishing knowledge management in schools of Dorud city using data mining techniques

Some models of confirmation of Knowledge management have used successfully in industrials, but as we know the features of educational organization such as a schools are different from industrials now, the question is that; is Possible to confirm the Knowledge management at schools?  At last, solving this time by Deans of schools is an important issue in Education organ.
This study is an applicable Survey, and is an analytical-descriptive in method to analysis the data used in the Data mining in this research. The related variants to predisposing factors of confirmation of Knowledge management are: the culture of organization and the process of Knowledge and information Technology the validity of process Control.
The questions are presented to some faculty members after editing, then Some of them which are included the problem were reviewed after reform supposing till the deletion of unsatisfied questions after that they were replaced by new others, so the external validity of the questionnaire was - Detected. at the reliability of process Control.
The studying of reliability of Test was held by Alfa Kron Batch methods and also by making halers. The reliability coefficient of the questionnaire is 92% which is a great one and by making halves we received to 70%, also is accepted. Finally, the questionnaire was planned with 35 questions in 5 multiples form which is said included three variables (the system of Knowledge processing,
The information technology system, and the cultural organization), then distributed between the managers.
Population 
The population of this survey includes the Deans of schools in Dorud’s education which are 116 people then selected 89 of them based on Morgan Table.
Using the algorithm of data mining and assessing the algorithms 
The decision tree, Baye's theorem and artificial Neural network were used in this survey, then had continued by confusion matrix which is one of the most important assessing features of casting -algorithm.
Decision tree
 Producing the rules using the decision tree.
After Inserting the preprocessing data in weka Software, then the rules below were produced by using the decision tree algorithm.
Rule1:  if the union accelerates less or lesser The Knowledge sharing for employees and the managers assent with their proposal more or less, it will be confirmed by the knowledge management.
Rule 2: if the union accelerated. Less or lesser the Knowledge sharing for employees and the manager’s assent with their proposals more and much more, it will be confirmed the knowledge management.
Rule 3: if the union accelerates more and more the knowledge sharing for employees and the managers would be less and lesser Sensitive to reserve their knowledge workers, then it will not be confirmed by the Knowledge management.

Rule 4: if the union accelerates more and more the knowledge Sharing for employees and the managers would be more and more sensitive to reserve their knowledge workers, then it will be confirmed the knowledge management.
The results were analyzed by using the efficiency features of the decision tree algorithm then to produce the below matrix by it.
The Confusion matrix of decision Tree





FP=1


TP=7




TN=10


FN=0





After executing the decision Tree, the concentration of casting would be 0.94.
The theory of Rough set function 
To extract the rules by this method using the Genetic, Johnson and Holts algorithms.
Genetic algorithm other produced 4100 rules, Johnson algorithm will be produced 18 rules and Holts algorithm will be produced 191 rules, and finally the most efficient rules will be presented among them.
Rule 1# if the managers are more Sensitive to reserve their Knowledge workers and the Union accelerates more or less the knowledge sharing for their employees then, it will be confirmed knowledge management. 
Rule#2: if the managers are almost sensitive to reserve their knowledge workers and the union accelerates more or less the Knowledge-sharing for their employees then, it will be confirmed knowledge management.

Rule 3# if the managers are less sensitive to reserve their knowledge workers and the union accelerates lesser the Knowledge-sharingfor their employees then, it will not be confirmed knowledge management.
The Confusion matrix of decision Tree





FP=1


TP=8




TN=2


FN=7





 
After executing the Rough Set Model, the concentrate of this algorithm would be 0.88.
 
According to the extracted rules from Genetic, Johnson, and Holts algorithm in Rough set model and decision Tree to receive the General and
Creditable rules from these four rules are Compound and finally made rules below.
The final Rule 1: if the Union accelerates almost or more the knowledge-sharing for their employees and the managers are more and much more sensitive to reserve their knowledge workers, then it will be confirmed knowledge management.
The final rule 2: if the union accelerates less Knowledge Sharing for their employees and the managers assent more and much more with their proposals, then it will be confirmed knowledge management.
The final rules 3: if the managers are less sensitive to reserve their knowledge workers and the union accelerates much less Knowledge-sharing for their employees, then it will not be confirmed knowledge management.
 The final Rule 4:
If the managers are almost sensitive to reserve their knowledge workers and the union accelerates much less Knowledge sharing for their employees Then, it will not be confirmed knowledge management.
 
Final Rule 5: if the managers are more sensitive to reserve their Knowledge workers and the union -accelerates almost the knowledge Sharing for their employees then it will be confirmed knowledge management.
Final Rule 6# if the managers are almost sensitive to reserve their Knowledge workers and the union accelerates much more knowledge sharing for their employees, and then it will be confirmed Knowledge management. 
Generally, according to the final findings in this survey which are held by data mining algorithm, the key features have an important role in conforming the knowledge a management are: Knowledge sharing, reserving the knowledge workers and using the employee's proposals.