Scatter/Gather Browser - a Visual Web Interface for Text Navigation and Search
Introduction
Specification
 Program List

Team members:
Alex Berry
Sujit Gadkari
Weimao Ke
Following the link below to the working Scatter/Gather Browser:
https://ella.slis.indiana.edu/g/l548s07c/browser/index.cgi
Following the link below to the working Scatter/Gather Browser:
https://ella.slis.indiana.edu/g/l548s07c/browser/help.html

Common Modules

inc.pm

This includes most common modules as follow: database connection and SQL statement preparation (prepareSQL), cluster distance measurement (clusterDistance), online fast hierarchical reclustering (cutTree), cluster visualization placement algorithm (placement), etc. This is to be included in all the cgi scripts in this project.

Node.pm

This is a module using Object-Oriented techniques in Perl to construct a holder of Node information. Each Node represents a cluster and is to be used for visualization. The basic information of a node includes TITLE, COLOR, X, Y, RADIUS, ID, DESCRIPTION.

project.js

Common javascripts for node visualization, placement, etc.

project.css

Cascading Style Sheet for the project, especially for the visualization.

 

CGI Programs

index.cgi

This is the primary interface of the scatter/gather browser. It captures user selections (i.e. clusters, # clusters to be reproduced), gather and rescatter the clusters to produce the new ones, and visualizes the new clusters on a 2D panel. When user mouses over a visualized cluster (i.e. node), it shows related information about the clusters--keywords and the centroid document. It also shows documents that belong to the selected clusters in a seperate frame functioned by "doc_list.cgi".

cluster_list.cgi

This script retrieves a list of clusters given
- selected parent clusters
- the number of clusters requested.
- the number of terms for cluster representation

cluster_show.cgi

This script retrieves detailed information about a cluster given its id (tree_id).

doc_list.cgi

This scripts retrieves a list of document records from the database
- given a the related clusters selected.

doc_show.cgi

This script retrieves a document record from the database given its doc_id.

 

Database

Schema

Data (big file)

 

Acknowledgements

Slider

We borrowed the Slider html/javascript codes from WebFX:
http://webfx.eae.net/dhtml/slider/slider.html
We appreciate their nice work on this.

(c) Copyright 2007 Scatter/gather -- Powered by Perl