Reference no: EM132384243

Visual Clustering Analysis of Social Network

This assignment gives students two options with different requirements based on student's disciplinary background, personal preference and existing experience. This is to satisfy the students who are not major in IT and Mathematics (Algorithms).

Option One: (group work):

A group of two students are required to work together to analyze an organization's email network (a type of social networks) through the data clustering and a clustered graph visualization.

Task 1: through data clustering, we can identify abnormal (implicit) network patterns that against the hierarchical structure of the organization,

Task 2: through a clustered graph visualization, we can visually read and quickly understand the data clustering output, including the abnormal network patterns.

Option Two: (individual work):

An individual student is required to visualize an organization's email network (a type of social networks) through graph visualization and clustered graph visualization.

Task 1: using graph visualization to visualize attributed email network,

Task 2: using clustered graph visualization to visualize a given clustered email network that enables readers to quickly understand the data clustering output.

General Requirement:

Option One: (group work)

Students are required:

1) To draw (visualize) the original email network on the paper (or screen) with the satisfaction of the following Aesthetics Rules: a) Symmetrical Display, b) Minimization of Edge-Crossings and c) Maximization of Angular Resolution. In addition, since each edge e in the graph is associated with a weight w(e), you need to map the w(e) to a graphical attribute, such as color, types of line, size or shapes, to enhance the readbility of the weight

2) To cluster this email network (or graph) into clustered structures by using Markov Clustering Algorithm. You need to produce two clustered structures 1) with the weight w(e), 2) without the weight w(e) .

3) Discuss the findings. If there is one (or more) abnormal network pattern(s) found, you need to describe them in details.

4) To draw (visualize) these two clustered graphs (one with w(e), another without w(e) ) on the paper (or screen). Using geometric rectangles (or circles) to bound clusters in the drawing. Make sure that these regions are not overlapped. In addition, these drawings shall also satisfy the general graph drawing aesthetics.

Option Two: (individual work) Student is required:

1) To draw (visualize) the original email network on the paper (or screen) with the satisfaction of the following Aesthetics Rules: a) Symmetrical Display, b) Minimization of Edge-Crossings and c) Maximization of Angular Resolution. In addition, since each edge e in the graph is associated with a weight w(e), you need to map the w(e) to a graphical attribute, such as color, types of line, size or shapes, to enhance the readability of the weight

2) To draw (visualize) a given clustering of the above email graph, that is: {0, 1, 2}, {3, 4, 5}, {6, 7, 8}, {9, 10, 11}, {12, 13, 14}, {15, 16, 17} on the paper (or screen).

Using geometric rectangles (or circles) to bound clusters in the drawing. Make sure that these regions are not overlapped. In addition, these drawings shall also satisfy the general graph drawing aesthetics, including a) Symmetrical Display, b) Minimization of Edge-Crossings and c) Maximization of Angular Resolution.

Attachment:- Visual Clustering Analysis of Social Network.rar