© 2002, 2003, 2004 Callipygian Show Systems, LLC. All rights reserved
Callipygian Grab 2.0, User’s Guide
Please note:
Callipygian Grab 2.0 is currently in the experimental stage, demonstrating the
use of Bayesian classifiers and neural networks to learn what a relevant web
site is. The code base changes often and the current state of the project may
vary slightly from the descriptions and images seen here.

Figure 1: Program running with 4 images found so far...
· Overview
o .NET Runtime with latest patches
o Windows XP Service Pack 2
o Lots of RAM and Disk space
o Types of searches
§ Browser History
§ File
§ Direct URL Entry
o Recognition of “Link Sites” and “Search Engine” sites
o Resume previously saved search
· Automatic Site-type classifier
o Normal Site
o Link Site
o Search Engine Result Page
o In HTML
o In PDF
o In JavaScript
o In Flash/Shockwave files
· Global Default Options
o Minimum/Maximum Image Dimension (with early-out detection)
o Minimum/Maximum Image file size (with early-out detection)
o Don’t Leave Site
o No Parent
o Max numbers of files per site
· Cache options
o Only download files checked more than n days/hours ago
o Clear Cache
o How Callipygian 2.0 learns what you’re looking for when you review photos
§ Bayesian Options
§ Bayesian Dictionary Review
§ Neural Network Options
· Image Review
o Crop / Rotate images before saving
o Discarding images
o Clearing image cache
· Minimize Mode view
o Using the thumbnail widget

Figure 2: Application in "Minimized" mode
· Report Generation
o Generating HTML-formatted website from search results
o Using XSL style sheets to “skin” the result page
Questions? Comments? Try the CallipygianGrab 2.0X Blog
©2004 Callipygian Show Systems, LLC