Showing papers by "Michael J. Maher published in 2014"

A Novel Approach for Network Traffic Summarization

Abstract: Network traffic analysis is a process to infer patterns in communication. Reliance on computer network and increasing connectivity of these networks makes it a challenging task for the network managers to understand the nature of the traffic that is carried in their network. However, it is an important data analysis task, given the amount of network traffic generated. Summarization is a key data mining concept, which is considered as a solution for creating concise yet accurate summary of network traffic. In this paper, we propose a new definition of summary for network traffic which outperforms the existing state-of-the-art summarization techniques. Our approach is based on clustering algorithm which reduces the information loss incurred by the existing techniques. By analysing the traffic summarization results using most up to date evaluation metrics, we demonstrate that our approach achieves better summaries than others on benchmark KDD cup 1999 dataset and also on real life network traffic including simulated attacks.

Complexity of Exploiting Privacy Violations in Strategic Argumentation

Abstract: Recently, Governatori et al. (2014) formulated a notion of strategic argumentation in the context of a dialogue game with incomplete knowledge, where arguments are constructed in a defeasible logic. Such a framework reflects aspects of the practice of legal argumentation. They show that a certain element of reasoning within strategic argumentation is NP-complete. In this paper we establish several related complexity results. To begin with, we present a much simpler proof of this result. Then the result is extended to allow the players the flexibility to have a wider variety of aims, and to address reasoning in a broad range of defeasible logics. Finally, we introduce some computational problems arising from violation of the privacy of a player in a strategic argumentation game, and establish their complexity.

Heart Disease Diagnosis Using Co-clustering

Abstract: Due to the advancement of information technology and its incorporation in various health applications, a huge amount of medical data is being produced continuously. Consequently, efficient techniques are required to analyse such large datasets and extract meaningful information as well as knowledge. Disease diagnosis is an important application domain of data mining techniques and can be resembled with the anomaly detection which is one of the primary tasks of data mining research. In past decades, heart disease caused the maximum death all over the world. As a result, heart disease diagnosis is a challenge for both data mining and health care communities. In this paper, co-clustering is introduced as a powerful data analysis tool to diagnose heart disease and extract the underlying data pattern of the datasets. The performance of the proposed method is evaluated using Cleveland Clinic Foundation Heart Disease dataset against other existing clustering based anomaly detection techniques. Experimental results reflect not only better accuracy but also meaningful information about the dataset which is helpful for further analysis of heart disease diagnosis.

Comparing defeasible logics

Abstract: In this paper we seek to formally establish the similarities and differences between two formalizations of defeasible reasoning: the defeasible logics of Nute and Maier, and defeasible logics in the framework of Antoniou et al. Both families of logics have developed from earlier logics of Nute, but their development has followed different paths and they are formulated very differently. We examine these logics from the standpoint of relative inference strength - how much the logics can infer from a given theory - and relative expressiveness - how well one logic can simulate another. We identify similarities between logics in the two families and pinpoint aspects that distinguish them.

An Efficient Approach for Complex Data Summarization Using Multiview Clustering

Abstract: There is significant interest in the data mining and network management communities to efficiently analyse huge amount of network traffic, given the amount of network traffic generated even in small networks. Summarization is a primary data mining task for generating a concise yet informative summary of the given data and it is a research challenge to create summary from network traffic data. Existing summarization techniques are based on clustering and frequent itemset mining which lacks the ability to create summary for further data mining tasks such as anomaly detection. Additionally, for complex and high dimensional network traffic dataset, there is often no single clustering solution that explains the structure of the given data. In this paper, we investigate the use of multiview clustering to create meaningful summary from network traffic data in an efficient manner. We develop a mathematically sound approach to select the summary size using a sampling technique. The main contribution of this paper is to propose a summarization technique for use in anomaly detection. Additionally, we also propose a new metric to evaluate summary based on the presence of normal and anomalous data instances. We validate our proposed approach using the benchmark network traffic dataset.

Strategic argumentation under grounded semantics is NP-complete

Abstract: We study the complexity of the Strategic Argumentation Problem for 2-player dialogue games where a player should decide what move to play at each turn in order to prove (disprove) a given claim. We shall prove that this is an NP-complete problem. The result covers one the most popular argumentation semantics proposed by Dung [4]: the grounded semantics.