Puteri N. E. Nohuddin

Trend mining in social networks: a study using a large cattle movement database

This paper reports on a mechanism to identify temporal spatial trends in social networks. The trends of interest are defined in terms of the occurrence frequency of time stamped patterns across social network data. The paper proposes a technique for identifying such trends founded on the Frequent Pattern Mining paradigm. The challenge of this technique is that, given appropriate conditions, many trends may be produced; and consequently the analysis of the end result is inhibited. To assist in the analysis, a Self Organising Map (SOM) based approach, to visualizing the outcomes, is proposed. The focus for the work is the social network represented by the UKs cattle movement data base. However, the proposed solution is equally applicable to other large social networks.

Detecting Temporal Pattern and Cluster Changes in Social Networks: A Study Focusing UK Cattle Movement Database

Temporal Data Mining is directed at the identification of knowledge that has some temporal dimension. This paper reports on work conducted to identify temporal frequent patterns in social network data. The focus for the work is the cattle movement database in operation in Great Britain, which can be interpreted as a social network with additional spatial and temporal information. The paper firstly proposes a trend mining framework for identifying frequent pattern trends. Experiments using this framework demonstrate that in many cases a large number of patterns may be produced, and consequently the analysis of the end result is inhibited. To assist in the analysis of the identified trends this paper secondly proposes a trend clustering approach, founded on the concept of Self Organizing Maps (SOMs), to group similar trends and to compare such groups. A distance function is used to compare and analyze the changes in clusters with respect to time.

Visualisation of trend pattern migrations in social networks

In data mining process, visualisations assist the process of exploring data before modeling and exemplify the discovered knowledge into a meaningful representation. Visualisation tools are particularly useful for detecting patterns found in only small areas of the overall data. In this paper, we described a technique for discovering and presenting frequent pattern migrations in temporal social network data. The migrations are identified using the concept of a Migration Matrix and presented using a visualisation tool. The technique has been built into the Pattern Migration Identification and Visualisation (PMIV) framework which is designed to operate using trend clusters which have been extracted from big network data using a Self Organising Map technique. The PMIV is also aimed to detect changes in the characteristics of trend clusters and the existence of communities of trend clusters.

Trend cluster analysis using self organizing maps

Trend cluster analysis using Self Organization Maps (SOM) is an application for clustering time series data. The application is able to cluster and display the time series data into trend lines graphs, and also identify trend variations in time series data. The system can process a large number of records as well as a smaller datasets. The results generated by the application are useful for analyzing large data which is often hard to analyze using normal spreadsheet software. The system has been developed using Matlab SOM functions and adopted SOM learning technique to cluster time series data. Based on the experiments, the test results have shown that the application is able to accommodate large sets of data and produce the trend lines graphs.

Trend Mining and Visualisation in Social Networks

A framework, the IGCV (Identification, Grouping, Clustering and Visualisation) framework, is described to support the temporal analysis of social network data. More specifically the identification and visualisation of “traffic movement” of patterns in such networks, and how such patterns change over time. A full description of the operation of IGCV is presented, together with an evaluation of its operation using a cattle movement network.

Frequent pattern trend analysis in social networks

This paper describes an approach to identifying and comparing frequent pattern trends in social networks. A frequent pattern trend is defined as a sequence of time-stamped occurrence (support) values for specific frequent patterns that exist in the data. The trends are generated according to epochs. Therefore, trend changes across a sequence epochs can be identified. In many cases, a great many trends are identified and difficult to interpret the result. With a combination of constraints, placed on the frequent patterns, and clustering and cluster analysis techniques, it is argued that analysis of the result is enhanced. Clustering technique uses a Self Organising Map approach to produce a sequence of maps, one per epoch. These maps can then be compared and the movement of trends identified. This Frequent Pattern Trend Mining framework has been evaluated using two non-standard types of social networks, the cattle movement network and the insurance quote network.

Social Network Trend Analysis Using Frequent Pattern Mining and Self Organizing Maps

A technique for identifying, grouping and analysing trends in social networks is described. The trends of interest are defined in terms of sequences of support values for specific patterns that appear across a given social network. The trends are grouped using a SOM technique so that similar tends are clustered together. A cluster analysis technique is then applied to identify “interesting” trends. The focus of the paper is the Cattle Tracing System (CTS) database in operation in Great Britain, and this is therefore the focus of the evaluation. However, to illustrate the wider applicability of the trend mining technique, experiments using a more standard, car insurance, temporal database are also described.

The application of social network mining to cattle movement analysis: introducing the predictive trend mining framework

This paper describes a predictive social network mining framework which is demonstrated using the Great Britain cattle movement datasets. The proposed framework, the predictive trend mining framework (PTMF), is used to analyse episodes of time-stamped social network data. The PTMF has two main components (1) a frequent pattern trend analysis component that efficiently identifies temporal frequent patterns and trends and also provides a mechanism for clustering and analysing these patterns and trends so as to detect dynamic changes within the cattle movement network, and (2) the predictive modelling component for forecasting the percolation of information or data across the network. The PTMF incorporates a number of novel elements including mechanisms to: (1) identify temporal frequent patterns and trends, (2) cluster large sets of trends, (3) analyse temporal clusters for pattern trend change detection, (4) visualise these changes using pattern migration network maps and (5) predict the paths whereby information moves across the network over time.

VisualUrText: A Text Analytics Tool for Unstructured Textual Data

The growing amount of unstructured text over Internet is tremendous. Text repositories come from Web 2.0, business intelligence and social networking applications. It is also believed that 80-90% of future growth data is available in the form of unstructured text databases that may potentially contain interesting patterns and trends. Text Mining is well known technique for discovering interesting patterns and trends which are non-trivial knowledge from massive unstructured text data. Text Mining covers multidisciplinary fields involving information retrieval (IR), text analysis, natural language processing (NLP), data mining, machine learning statistics and computational linguistics. This paper discusses the development of text analytics tool that is proficient in extracting, processing, analyzing the unstructured text data and visualizing cleaned text data into multiple forms such as Document Term Matrix (DTM), Frequency Graph, Network Analysis Graph, Word Cloud and Dendogram. This tool, VisualUrText, is developed to assist students and researchers for extracting interesting patterns and trends in document analyses.

Analyzing and Detecting Network Intrusion Behavior Using Packet Capture

Network Intrusion is one of serious computer network security issues faced by almost all organizations or industries around the world. The big problem is that companies still have poor security to keep their network in good condition. Unfortunately, the management takes the simplest way by putting heavy responsibilities to network administrator rather than spending a high cost of computer security setup. In this paper describes a preliminary study for proposing a technique of analyzing network intrusion by using Packet Capture integrated with Network Intrusion Behavior Analysis Engine. This technique analyzes whether the flow of the network is healthy or malicious. The study consists of several components for implementing an effective and efficient network analyzing mechanism. Artificial Neural Network is selected as the main method for its behavior analysis engine. Then, it will illustrate the analysis result using an enhanced visualization method which gives more knowledge and understanding to the network administrators for effectively monitor network traffics.