Browsing by Author "Bhuiyan, Md Hasanuzzaman"

Now showing 1 - 3 of 3
  • Thumbnail Image
    Large Scale Network Visualization with Gephi
    Alam, Maksudul; Arifuzzaman, S. M.; Bhuiyan, Md Hasanuzzaman (2012-12-11)
    The notion of graphs or networks is sufficiently pervasive since it can be used to model various types of data sources. Social, biological, and other networks capture the underlying structural and relational properties. Analysis of different networks reveals interesting information of the corresponding domain or system. Network analysts, thus, strive to analyze various networks by applying different algorithms and try to connect obtained insights to make sense of a unified theme, pattern or structure. For example, analysis of facebook friend network of a person can reveal information such as, groups of highly clustered people, most influential person in terms of connections, connecting persons between different cluster of people, etc. While analyzing networks and digesting the information therein, analysts gradually form internal mental models of the people, places, events, or any sort of entity represented in the networks. As the number of nodes grows larger, however, it becomes increasingly difficult for an investigator to track the connections between data and make sense of it all. Many researchers believe that visual representations of data can help people better examine, analyze, and understand them. Norman [Norman94] has described how visual representations can help augment people’s thinking and analysis processes. The objective of the project is to develop visual representations of nodes, edges, and labels of a network in order to help analysts search, review, and understand the network better. We seek to create interactive visualizations that will highlight and identify significance of nodes, cluster formation, etc., in the networks where entities may be, for example, people, places, webpage, biological entity, dates and organizations. Basically, we want to build visual representations of the networks that help analysts making sense by applying different algorithms on them and observe the difference of nodes and edges in terms of color, and size. A very important aspect of the project is the integration of the visualization module with CINET [CINET2012], a cyberinfrastructure for network science. CINET includes a set of graph algorithms and various types of networks. Analysis of networks are done by applying algorithms on those networks; results are obtained as text files containing information of different measures of nodes or edges. Complex workflow is intended while working with CINET where output of one analysis can be used as input to further analysis. Visualization comes as a great aid when analyst want to filter his interest on some particular nodes or a portion of the graph and conduct subsequent analysis on the smaller part. Though there are some existing visualization tools, e.g., Jigsaw [Jigsaw08], Sentinel Visualizer, NetLens, etc., they are more focused on information representation rather than on graph exploration or summarization capabilities. To the best of our knowledge, our project is the only one which supports network visualization as a part of complex workflow within network analysis utilizing high performance computing environment. In summary, this project develops a visualization component for a VT digital library containing large network graphs (e.g., social networks and transportation networks). The visualization service will get datasets from an existing DL, visualize the graphs using Gephi (a java-based visualization library), and integrate the results within an NSF supported cyberinfrastructure (CINET).
  • Thumbnail Image
    Parallel Algorithms for Switching Edges and Generating Random Graphs from Given Degree Sequences using HPC Platforms
    Bhuiyan, Md Hasanuzzaman (Virginia Tech, 2017-11-09)
    Networks (or graphs) are an effective abstraction for representing many real-world complex systems. Analyzing various structural properties of and dynamics on such networks reveal valuable insights about the behavior of such systems. In today's data-rich world, we are deluged by the massive amount of heterogeneous data from various sources, such as the web, infrastructure, and online social media. Analyzing this huge amount of data may take a prohibitively long time and even may not fit into the main memory of a single processing unit, thus motivating the necessity of efficient parallel algorithms in various high-performance computing (HPC) platforms. In this dissertation, we present distributed and shared memory parallel algorithms for some important network analytic problems. First, we present distributed memory parallel algorithms for switching edges in a network. Edge switch is an operation on a network, where two edges are selected randomly, and one of their end vertices are swapped with each other. This operation is repeated either a given number of times or until a specified criterion is satisfied. It has diverse real-world applications such as in generating simple random networks with a given degree sequence and in modeling and studying various dynamic networks. One of the steps in our edge switch algorithm requires generating multinomial random variables in parallel. We also present the first non-trivial parallel algorithm for generating multinomial random variables. Next, we present efficient algorithms for assortative edge switch in a labeled network. Assuming each vertex has a label, an assortative edge switch operation imposes an extra constraint, i.e., two edges are randomly selected and one of their end vertices are swapped with each other if the labels of the end vertices of the edges remain the same as before. It can be used to study the effect of the network structural properties on dynamics over a network. Although the problem of assortative edge switch seems to be similar to that of (regular) edge switch, the constraint on the vertex labels in assortative edge switch leads to a new difficulty, which needs to be addressed by an entirely new algorithmic approach. We first present a novel sequential algorithm for assortative edge switch; then we present an efficient distributed memory parallel algorithm based on our sequential algorithm. Finally, we present efficient shared memory parallel algorithms for generating random networks with exact given degree sequence using a direct graph construction method, which involves computing a candidate list for creating an edge incident on a vertex using the Erdos-Gallai characterization and then randomly creating the edges from the candidates.
  • Thumbnail Image
    Text Classification Using Mahout
    Alam, Maksudul; Arifuzzaman, S. M.; Bhuiyan, Md Hasanuzzaman (2012-11-06)
    This module focuses on classification of text using Apache Mahout. After successful completion of this module, students will be able to explain and apply methods of classification, correctly classify a set of documents using Apache Mahout, and construct and apply workflows for text classification using Apache Mahout.