A technique for mining association rules in multi dimensional data bases

مقدم أطروحة جامعية

Hasan, Shayma Akram

الجامعة

الجامعة التكنولوجية

الكلية

-

القسم الأكاديمي

قسم علوم الحاسوب

دولة الجامعة

العراق

الدرجة العلمية

ماجستير

تاريخ الدرجة العلمية

2008

الملخص الإنجليزي

Data mining is the process of extracting knowledge or patterns of data hidden from large volumes of raw data.

The knowledge must be new, not obvious, and one must be able to use it.

Data mining finds these patterns and relationships using data analysis tools and techniques to build models.

One of the most important data mining techniques is the association rules, It studies the frequency of items occurring together in databases based on support and confidence.

An attempt to improve the efficiency of the association rules technique in data mining is done by proposing an algorithm that deals with multidimensional databases with extremely long item sets.

This algorithm focuses on the long size of item sets while the size of database is not taken in the account is called ARMDLI (Association Rules Mining in Multidimensional Databases with Long Item sets.

It divides the multidimensional database to two parts : the Dimension's Part of the Multidimensional database (DPM) and the Itemsefs Part of the Multidimensional database (IPM) then mine each part separately.

In the dimension's part, two methods (algorithms) for finding the set of frequent Litems in that part have been proposed to overcome inefficient direct access of the traditional algorithms.

The association rules here which consist of positive ; and negative rules have been extracted using a modified traditional association pules procedure, In the itcmset's part, the set of the most frequent itemsets has large search space in which it is important to speed up the process and save the storage area.

The remaining frequent itemsets are searched and appended to the previous set.

The complete set of frequent itemsets has been classified to normal, closed and maximal to support and strengthen the prediction in the analysis stage.

And then the set of association rules for the itemset's part has been extracted using the traditional association rules procedure.

The proposed system has been applied to a proposed web log database to extract the association rules from it instead of the traditional methods.

التخصصات الرئيسية

تكنولوجيا المعلومات وعلم الحاسوب

الموضوعات

• الإدارة
• قواعد البيانات
• التنقيب في البيانات

• APA
• MLA
• AMA

لغة النص

الإنجليزية

نوع البيانات

رسائل جامعية

رقم السجل

BIM-305499