The magazine of the Melbourne PC User Group
SearchPad 1.1
Bob Burt
 bobburt@melbpc.org.au

Among other things, the Internet is an ever-expanding mass of Web sites and data storage resources, which the visitor must learn to navigate sensibly and effectively to feed his or her particular interests. Many sites designed to service particular topics provide links to others of similar content and there is a very useful handful of search engines that you can employ to further expand such lists. Some sites even group the search engines for you and provide the means to streamline such searches. For example, I occasionally use one which will identify any MIDI files available which meet particular parameters I have nominated.

SearchPad goes a very useful step or two further than this. Essentially, it is a self-standing utility designed to search for, gather, organise, analyse and present online information that meets the descriptions you have asked for.

The package is available on CD-ROM and is supported by a compact 16-page manual. If the utility interests you, you can test drive a copy available as an .EXE file (version 1.1 R2) from http://www.searchpad.com/.

 


Figure 1. SearchPad displaying online information


Figure 2. The Query Wizard - setting keyphrases

Installing SearchPad

An Install Shield Wizard guides you through installation and provides a default destination drive and directory which you can modify if you wish.

Default preferences are set automatically to

  • 10 hits per search engine
  • A scan depth of 1 for documents
  • A cache size of 10 MB
  • A file size limit of 10 KB.
Using SearchPad

As a new user, you will find the Query Wizard most helpful, as it automates the search process. Essentially, you first set up a Query, which is a word or phrase given to trigger the search. Follow the illustrations to see the sequences you need to set up.

Along the way, you nominate which search engines you wish to use, selected from OneKey, Altavista_Usenet, HotBot, OpenText, Magellan, Excite, Infoseek, Altavista, Yahoo, Lycos and Webcrawler.


Figure 3. Selecting search engines


Figure 4. Naming the topic

As an exercise to generate illustrations for this article, I nominated a subject of Jazz Piano and used the query Jimmy Yancey, simply because he was an important developer of a particular jazz style but is not very well-known, so information is likely to be rather sparse, even on the Internet. Using three search engines - Altavista, Yahoo and Lycos - SearchPad retrieved 50 titles of URLs, although only the first 26 of these were fully relevant to the query. This means that the remaining 24 referred to either Jazz Piano or just Jazz, without any mention of Jimmy Yancey. Each of these entries is annotated "Does not conform to query".

A useful feature of the retrieval operation is that duplicated URLs found by different search engines are automatically filtered out, although multiple URLs can be present which are linked to each other in the Web pages.

The retrieved information is displayed very neatly on the screen and care has been taken to provide a printout with a good layout and pleasing choice of fonts.

You can frame Rules for both Queries and Topics and you may have multiple queries in relation to particular topics. You normally design query rules before topic rules. Rules are framed by using keyphrases and the format includes Boolean reserved words such as IN, NOTIN or AND. It sounds complicated, but the manual includes some simple examples for you.

 


Figure 5. Adding a crawl URL


Figure 6. Summary of search details

The Query dialog box provides a Crawl URLs tab which allows you to nominate a set of URLs to be downloaded together with their links. The document is then analysed and classified. The crawling depth to which the document is scanned is set in the General Preferences dialog box.

Feedback is available as a secondary level of filters (the primary level being the filters that Rules provide). Again, the manual gives a useful example. You can view documents from any category and provide feedback on them.

 


Figure 7. The Query dialog box


Figure 8. Displaying a URL and classifying it

After you have completed a search and have the URLs listed on the screen, you double-click on each of these in turn to call up a new screen to rate them as highly relevant, relevant, average, irrelevant or highly irrelevant. You can also add synonyms and provide feedback, if available. If you are still on the Internet while you are doing this, you can click on any of the links provided in the URL window and connect to them.

Conclusion

This is a very neat utility that will become ever more useful as you continue to use it, as it becomes smarter and more effective with repeated use.

Reprinted from the May 1998 issue of PC Update, the magazine of Melbourne PC User Group, Australia

[About Melbourne PC User Group]