كشف الفيروسات باستخدام تقنيات التنقيب عن المعطيات

الموضوعات

علوم تطبيقية وتكنولوجية

روابط وملفات

التعريف والنوع

رقم الوثيقة

0456eb03-f298-4aea-8526-cd08292e79d1

رقم العقد

0

نوع الوسائط

Crawler

نوع المحتوى

الرسائل العلمية

صيغة المصدر

رسائل دكتوراة

نوع الملف

pdf text

أسماء الملفات

1958860_1.pdf

بيانات النشر

ترجمة العنوان

Virus Detection Using Data Mining Technologies

ألقاب المؤلفين

[{"name_ar":"عصام قزموز","title_ar":"اعداد","title_en":"Preparation"},{"name_ar":"يحيي نجار","title_ar":"اشراف","title_en":"Supervision"}]

اللغة

Arabic

المصدر والدورية

اسم المصدر

كشف الفيروسات باستخدام تقنيات التنقيب عن المعطيات

المحتوى والصفحات

عدد الصفحات

136

ترجمة الملخص

In this thesis we introduced the idea about viruses and the differences between the classic virus detection and virus detection using data mining. Where the classic method can detect the viruses but they need to know the virus and execute it before they can detect the virus because they depend on Signature of Virus that represent of number of bytes sequence. For that it is very important to find a new method that can help us to detect the new virus without execute them.Byte sequence of the virus programs and benign programs gives a lot numbers of byte especially by using bytes sequences n-gram. For this reason it needs feature extraction methods to choose the best features as information gain and class-wise method then enter these features into classification algorithms.We proposed a two method for features extraction that give a good results of feature extraction than class-wise and information gain methods. The first method depend on the number of times that this feature appears on each file and how many times it repeated whereas the class-wise method depend on the number of times it appears in the files.The second method depend on the H standard measure for selecting the best feature by selecting the features obtained the big H value.The proposed two methods give best detection rate and minimize the false alarm rate.The chosen features are used as an input for classification algorithm to check if the file has a virus or benign.Pruning method for RULES-5 has been proposed.This proposed pruning method made the rules of the RULES-5 more general and the number of rules have been reduced comparing with RULES-5.

إشراف وإعداد

الإشراف

يحيي نجار

الإعداد

عصام قزموز

الاقتباسات الببليوغرافية