What is twURL™?
In a nutshell, twURL™ is a powerful Windows 95 Web
utility for viewing and managing collections of URLs extracted from downloaded
HTML files.
WWW Information Professionals can use twURL many ways,
in concert with other web utilities, to improve their web work productivity,
apply quality criteria to search results, and generate useful reports.
What can you do with twURL ?
Consider twURL as a visual list manipulator tuned for
WWW use. It lets you:
- See the outline structure of URL collections organized by:
- Host, e.g. www.whitehouse.gov, www.w3.org, www.digital.com, www.mit.edu,...
- Domain, e.g. gov, edu, com or international ones: ca, fi, ...
- Strings in URLs or link labels, number of links to, common links, site
sizes, etc.
- Vocabularies of terms and web pages containing them
- Your own subject hierarchies - like bookmarks, but much more you can
do with them
- Ratings applied during browsing
- Tag a URL with a "webit", a note with anything you want yourself
or future readers to know, as well as automatically extracted HTML Meta
fields and page prefaces
- "Report out" the twURL outlines into HTML files
with an outline appearance using bullets, with menus or frames for navigation.
- Navigate URLs the way you've organized them, like itineraries for a
browsing session (yes, twURL !is! a browser itself)
Why would you want to do these things?
Well, first of all, twURL makes it really easy to visualize
any URL collection -- just designate the directory and HTML or shortcut
files and you'll get a number of different ways to sort (or, as we say,
"twURL") the URLs in a new way. twURL can:
- "X-ray" the HTML document skeleton at various sites
- to see how deep you have to go to find what you want or to help find
related matter at a productive site, or find out where the most action
is on a topic, even showing up HTML document errors
- give you overviews, as well as specific categorization,
- of what material is at different sites you've visited (possibly in
many different search sessions, combining several search engines, and maybe
using other tools that download bunches of Web material at one time)
- help you more systematically organize results from previous searches
- into collections or pick a starting point for more searches (for example,
from an existing but untapped link)
- organize the results you get from other web utilities,
- and provide work for them, mobilizing more tools for your work
- help you track changes in URLs
- by replaying searches or using twURL in concert with
other monitoring tools on the market
- review your browsing history
- (using history or cached files), showing you where you've concentrated
your time and providing a "due diligence" trail
- plan your browsing sessions across 100s of URLs
-
- by downloading and previewing URLs - no more surfing blindly (unless
you want to)
- document like a professional researcher
-
- by attaching your own notes, editing automatically extracted prefaces
and meta-information, and applying your own rating/relevance scheme
Once you start using twURLed organizations of your HTML collections,
you'll begin to see lots of patterns in the URLs and expand your use of
valuable content scattered across the sites you've been using.
Can you give me an example of what twURL does?
Absolutely! In fact, we're trying to codify our experience in our twURLed
World article series.
- Suppose you're interested in finding some highly popular and well-regarded
sites on the web and you've just located and downloaded the top 100 lists
from several magazine and tracking sites. You need to compare several lists
(in their HTML forms) and determine which URLs they have in common. twURL
will build these lists, apply its "Twizards" to show you the
exact overlaps, and let you edit the resulting lists for your own purposes.
- Now consider yourself in the situation where you need to put together
an extensive collection of URLs for your boss (or your website) to lay
out what's available on the Web regarding the "Year 2000 Problem"
(when old code turns back the clock a century). Your favorite metasearcher,
WebFerret, hauls in over 5000 URLs from search engines. Being a pro-research,
long-term minded organization, you also want to build some bridges to the
research community and to take advantage of not just what's on the market
but also of promising new ideas. twURL will help you collect, organize,
filter, rate, HTML-ize, and browse URLs.
What's different about twURL?
Right now, there's no URL visualizer like twURL on the
Web market, so we're just beginning to learn what twURLed
patterns mean, especially for professional searchers looking for "power
tools" to extend their browser capabilities. We've already shown you
twURL's powerful way of matching lists and annotating search
results. twURL provides an alternative to traditional one-URL-at-a-time
forward-backward linking supported by today's browser behemoths. You can
download, edit, plan, browse, annotate/rate, download, edit, plan, browse,
annotate/rate, etc. then produce readable HTML reports for distribution
and archiving.
Our ROI objective is to use these patterns for establishing a "context"
to the search results (a concept that will mean more and more as you get
deeper into the Web of URLs that are growing each day).
twURL will work on any normal downloaded HTML files, e.g.
ones you've saved from Netscape, MOSAIC, or Internet Explorer, and some
of the other web utilities on the market. There are many "pathfinder"
files that provide good starting points for twURL, since
you can examine alternative presentations of a good URL selection. You'll
see that twURL has many potential applications to HTML-based
documents and can readily be used in combination with other off-line, monitoring,
and meta-search Web utilities.
twURL works at a different range than most current browsers
or browser utilities. It's not uncommon for us at ROI to come back from
a Web topic search with a collection of 1000+ URLs to organize and explore
further. Most browser and bookmark utilities work one URL at a time, although
they can handle these well at the moment with immediate browser access,
and that won't cut it for large numbers of URLs to revisit. Of course,
the search engines we're all tapping to mine the WWW hold millions of URLs.
twURL aims to offer good control over the ranges of 300-3000
URLs at a time, which is about the size of most topic searches we're seeing
today (although usually around 250 URLs are significant).
If you pick up some good site lists on the Web during an extensive search,
you'll certainly want to use twURL to give you different
perspectives on what you've picked up (especially what was a waste of time!).
You'll soon see that twURL's simple approach -- sorting URLs
offers myriad different ways of viewing your search results.
What can't you do with twURL ?
Right now, twURL
- doesn't help you populate a URL collection or track searches real-time
in your current browser
- lacks a graphical interface. This is coming soon using the Visio package
to generate imagemaps for HTML reports and, possibly, act as a twURL graphics
engine (or vice versa)
- isn't intended to do all the things you can do with other WWW utilities
such as WebWhacker, WebWatch/Surfbot,
SmartMarks, SqURL/WebSeeker,
NetAttache, WebCompass,
Zooworks, etc. We've noted a few ways
to use twURL as a component of a coming suite of such tools.
- doesn't summarize document content beyond exposing the hyperlink structures
and obvious HTML headings and stripped first 500 words. We're investigating
how to use it with full text indexers. The end result is still very informative
HTML reports that look like annotated bibliographies.
- doesn't pull together multiple searches into databases, which will
reach into the hundreds of thousands of URLs to be managed for more integrated
topic exploration over longer time periods.
- won't fully resolve all URLs. Browser files were meant to be seen and
not computed from, apparently, because there are lots of ways URLs get
lost when files are downloaded or left dangling when relative URLs are
used.
- only runs on Windows (95, 98, NT)
What does twURL look like?
Check out the screenshots in the slideshow.
So, how do I actually get twURL ?
Sorry, we can't distribute it widely right now. Our past trials and
distributions didn't yield us much feedback so we're holding off until
we find the right suite of tools to inter-operate with. However, we welcome
serious experimenters and will provide a downloadable trial version upon
request.
Contact Information
Research Outlet and Integration/ROI JV
Tel: 281-486-8480 (Houston)
E-mail: info@twurl.com
Website: http://www.twurl.com
Copyright © 1996, 1997 ROI JV
Last modified: January 27, 1997/November 20, 1998