In recent years, there has been great attention given to skyline queries that incorporate and provide more flexible query operators that return data items (skylines) which are not being dominated by other data items in all dimensions (attributes) of the database. Many variations in skyline techniques have been proposed in the literature. However, most of these techniques determine skylines by assuming that the values of all dimensions for every data item are available (complete). But this assumption is not always true particularly for large multidimensional database as some values may be missing (not applicable during the computation). In this paper, we proposed an efficient approach for processing skyline queries in incomplete database. The experimental results show that our proposed approach has significantly reduced the number of pairwise comparisons and the processing time in determining the skylines compared to the previous approaches.

No abstract

No abstract

Nowadays, malware incident is one of the most expensive damages caused by attackers. Malwares are caused different attacks, so considerations and implementations of malware defences for internal networks are important. In this papers, different techniques such as repacking, reverse engineering and hex editing for bypassing host-based Anti Virus (AV) signatures are illustrated, and the description and comparison of different channels and methods when malware might reach the host from outside the networks are demonstrated. After that, bypassing HTTP/SSL and SMTP malware defences as channels are discussed. Finally, as it is important to find and detect new and unknown malware before the malware gets in to the victims, a new malware detection technique base on honeynet systems is surveyed.

Intrusion detection system (IDS) is used to detect various kinds of attacks in interconnected network. Many machine learning methods have also been introduced by researcher recently to obtain high accuracy and detection rate. Unfortunately, a potential drawback of all those methods is the rate of false alarm. However, our proposed approach shows better results, by combining clustering (to identify groups of similarly behaved samples, i.e. malicious and non-malicious activity) and classification techniques (to classify all data into correct class categories). The approach, KM+1R, combines the k-means clustering with the OneR classification technique. The KDD Cup '99 set is used as a simulation dataset. The result shows that our proposed approach achieve a better accuracy and detection rate, particularly in reducing the false alarm.

No abstract

No abstract

No abstract