t

w

U

R

L

(™)

context

continuity

control

to
WWW
content




(Windows 95)

Today's powerful browse, search, download, index, monitor, and display tools all increase speed and ease of access to the WWW's valuable content. But it's still far too difficult to organize the reams of retrieved materials, profile their characteristics, edit URL indices, and capture and retain context. twURL provides greater control, continuity, and context for valuable content by:

• extracting the "hyperlink skeleton" of HTML materials and visualizing that skeleton in familiar outline forms
• profiling key characteristics - most linked URLs and sites, overlaps among sources (e.g. search engines), and distribution across domains (.com, .edu, etc.)
• editing these outlines by marking properties, trimming, and rearranging into subject hierarchies
• browsing (in context) regions of those edited and reorganized outlines
• recording ratings and annotations as results are selected and qualified.

1. define a purpose for a WWW expedition
2. build an HTML collection using metasearcher/search engine output, pathfinder pages, automated downloads, and plain old one-URL-at-a-time browsing and bookmarking
3. read the HTML files into twURL then display, query and explore the outline, trimming off-topic URLs
4. analyze (using twURL's 'twizards') the hyperlink skeleton for most promising URLs and most interesting web regions to lay out an efficient "itinerary"
5. invoke an embedded browser to follow that itinerary and track side trips, while also recording impressions using 'webit' annotations and a user-defined rating form for relevance, significance, and utility
6. reorganize URLs into subject and rating clusters, sections for reports, and subsets for monitoring
7. generate HTML reports for sharing across nets and save sessions for monitoring, integration, and future use.

Scenarios from the twURLed World

Teaching Internet skills and critical abilities

Professor CH believes that all students need Internet-savvy to use WWW as an extended technical library for college courses and life-long learning. A typical assignment by CH is: "update the references of chapter X with recent advances in the field, locations of influential technology vendors and service providers (career pointers, market knowledge), and one or more critical evaluations of this subject matter. Make your own judgments on significance and utility of the web pages you find. Turn in your twURL .twb file with URLs, annotations, and ratings." Students use twURL and other fetch-and-monitor tools, plus their own growing library of HTML notes, to identify the most promising web regions, e.g. by extracting and then browsing the hyperlink skeleton of a Virtual Library page within twURL. Submitted results are merged and provided on their classroom network, amazing many students at the variety of materials found, as well as classmate's divergence of opinion. Professor CH integrates each exercise into a growing bookweb, e.g. Issues of the Year 2000, that keeps the course up-to-date, while challenging the students to develop their critical abilities.

Making a list, checking it twice

The Cream of the Web (CoW), http://www.bovine.org, features a weekly comparison of the "top 100", "favorite 50", "coolest 40", etc. lists from other webzines, offering its visitors the "best of the best". The CoW website crew used to pour over these lists, checking off common URLs, visiting them one at a time to check URL accuracy, and cutting and pasting URLs, all within HTML editors. Now, they schedule agents to download the latest changes, feed the lists into twURL, run automatic twizards for common URLs, edit commentary and ratings onto each URL (all while browsing the changed and new pages) then rearrange the URL outlines to fit their HTML templates and style sheets. twURL handles the editing, guides the browsing, and the CoW crew concentrate on the qualities they want to feature. For an extended example of this scenario, visit the ROI website http://www.twurl.com/.

Technical Advice for an Engineering Standard

SG is a web adviser for a major engineering product evaluation organization. As standards are promulgated, it is imperative that the wide-spread engineering community get immediate access to the technical literature as well as the marketplace of professional courses, product announcements, and service advisories. When SG produces a web-based technical advice topic, starting from an initial 1000 or so URLs, twURL acts as an editor and relevance recorder. SG also uses twURL to calibrate search engines to find the most thoroughly indexed, relevant, up-to-date, best summarized materials, and to counterbalance over-promotion by one vendor or university. twURL's 'twizards' locate the most promising web route then its automatically download (often 25% of links are dead) the web pages for SG to browse, categorize, and qualify at the rate of about 1 a minute. SG culls the best 50 URLs and another 250 good ones, organized by category and then hyperlinked back to the text of the standard. Where life and death, e.g. cases of software-caused fatalities, is the end goal, the most authoritative and quality assuring methods are essential.