SOM visua : data clustering and visualization based on SOM and GHSOM

Joint Authors

Makki, Muhammad A.
Kahlut, Khalid M.

Source

The fifth International Conference on Engineering and Sustainability : is scheduled on November 4-5, 2014

Publisher

Islamic University of Gaza Faculty of Engineering

Publication Date

2014-11-30

Country of Publication

Palestine (Gaza Strip)

No. of Pages

10

Main Subjects

Engineering & Technology Sciences (Multidisciplinary)

English Abstract

Text in web pages is based on expert opinion of a large number of people including the views of authors.

These views are based on cultural or community aspects, which make extracting information from text very difficult.

Search in text usually finds text similarities between paragraphs in documents.This paper proposes a framework for data clustering and visualization called SOMvisua.

SOMvisua is based on a graph representation of data input for Self-Organizing Map (SOM) and Growing Hierarchically Self Organizing Map (GHSOM) algorithms.

In SOMvisua, sentences from an input article are represented as graph model instead of vector space model.

SOM and GHSOM clustering algorithms construct knowledge from this article.

SOMvisua provides a visual animation for eight famous graph algorithms execution with animation speed control.

It also presents six types of visualization.

For visualization of similarity lists, we use well-known methods that take a similarity list as input and according to the used similarity measure; an adjustable number of most similar sentences are arranged in visual form.

In addition, this paper presents a wide variety of text searching.

We conducted experiments on the SOMvisua using a large document dataset.

Then we compared the performance with that of hierarchal clustering with automated topology based SOM and GHSOM clustering to prove the superiority of SOMvisua.

Data Type

Conference Papers

Record ID

BIM-774819

• APA
• MLA
• AMA