Detection of Attacks Executed by Multiple Users - Dissertation Example

Some of these mass-users attacks are triggered by the big companies and manipulation of online items reputation can be hazardous for the customers. As a rule, well-known products are chosen for this type of attacks. For example, Amazon books, some hotels in travel sites, and a great number of digital content can be a fertile ground for mass-users attacks. Not only these vicious and hazardous attacks can be the greatest challenge of the electronic life, but still Netizens and other computer world dwellers are intimidated by mass-users attacks.

Under conditions of this type of attacks, fraudulent users implement their well planned strategies and manipulate reputation of numerous target products. To consider these attacks and the ways of dealing with them, it is possible to apply a defense scheme that (1) develops heterogeneous thresholds for developing protection against suspicious products and (2) analyzes focus items on the basis of correlation analysis among suspicious items. Real user data and simulation data should be correlated and on the basis of such kind of correlation it is relevant to identify potential mass-user attackers.

The given scheme shows the main advantages in finding out fraudulent users, recovering challenging errors in the systems, and reducing attacks related to normal products, sites etc.   The problem of attacks executed by multiple users is a complicated task and the modern researches are working in the name of these fraudulent groups identification. The main task of the modern researchers in this field is to apply advanced artificial intelligence and complex adaptive systems to stop, foresee and prevent distributed attacks of the network.

Multiple attackers work together and very often it is difficult to foresee and prevent this type of an attack. Data Reduction Techniques is one of the most convenient means of preventing this type of attacks. IDS approaches are nowadays limited in a proper identification of relevant information in high-speed network data streams. The appropriate analysis of IDS enables taking control over such type of attacks. There is a need to conduct a dynamic and real-time control over current attacks. Relevant information can be processed and it can serve as an input vector to IDS.

There is another challenge, which is a great variety of activity and processes occurring in the network environment to identify a subset of data that is very difficult for analysis. There is a suggestion to design an anomaly detection scheme for prevention of mass online attacks. This underlying scheme is based on several components integration: first is the importance of time-domain change detection, the second step is the importance of system-level visualization, the third are the selection of heterogeneous threshold and a conduct of a proper correlation analysis.

Therefore, we can claim that for prevention and protecting computer systems from mass-users attacks it is necessary to pay attention to the new philosophy. Currently existent schemes of attack prevention and protection are mainly based on homogenous correlation of items and the proposed scheme provides a much better performance in the process of malicious users' detection and reducing impacts on normal issues. Due to a wide range of activities and processes, the identification of fraud in a network environme