Previous section Navigate within course Next section
Course Overview
 - Slide for this section

Session 1

What is the Internet?

Internet vs other information resources

Finding a 'knowledge hub'

Search engines

- How they work
 - Comparing engines
- Using Top Three sites

Search strategies

Offline Assignment 1

Session 2

Evaluating information
Researching companies
Researching markets
Researching countries
Researching news
Search software
Offline Assignment 2

Using Search Software

All the Internet-based general search engines are superb tools to help you manage the millions of documents on the World Wide Web, but there comes a time when you just need more: more control of the information you're retrieving, more options to save the that information; just more - and that's where Intelligent Agents come in.

Intelligent Agents (also known as 'bots') and other specialised search software are what make it possible for you to really start harnessing the power of the Web. There are many software packages out there - some free - that give you that extra bit of power to do your searching. And not just that: these programs will often offer you the option of saving whatever you find as a complete offline Web site, or they will write out information to a database.

Intelligent Agents (they should really be called 'Smart Agents' because they are not truly 'intelligent') are specialised utilities that can be programmed by the user to search the Web in a particular way. Using either their own spider, or making use of a suite of standard search engines, they crawl the Web looking information using the guidelines you have given them.

For one of these agents to behave in an 'intelligent' way, it must be told:

  • What to look for
  • Where to look for it

Telling An Agent What To Look For

This is usually done by means of a query string. Such a string is much like the kind of query string you pass to a Web search engine. It is a collection of terms that describe the kind of data you are looking for with a standard Boolean format. This query string acts as a filter for the data retrieved by the agent.

Telling An Agent Where To Search

This is the critical part of setting up an agent. You must tell it where on the Web to look for information. In the case of an agent with its own spider, this can be any Web site you can think of; in the case of an agent that uses other search engines, this consists of telling the agent which of those engines to use. This careful selection of resources to search acts as another filter on the data retrieved by the agent.

Examples Of Intelligent Agents And Bots

There are literally hundreds of agents out there, and many different types. Some are Web-based and some are stand-alone applications.

Here are a few examples:

  • Data Mining Bots, e.g. MOMspider - Multi-owner Maintenance Spider (MOMspider) is a web-roaming robot that specializes in the maintenance of wide area webs.
  • E-Mail Bots, e.g. SmartBot - the intelligent autoresponder featuring a complete site offering tests, laws, instant sign up, articles and documentation.
  • Government Bots, e.g. The Data FERRET (Federal Electronic Research and Review Extraction Tool) - is a computer search tool developed by the Census Bureau and the Bureau of Labor Statistics.
  • Knowledge Bots, e.g. AgentBuilder - an integrated tool suite for constructing intelligent software agents.
  • Commerce Bots, e.g. AgentBuilder for eCommerce - AgentBuilder is an integrated toolkit that enables you to create intelligent agents quickly and easily. This application consists of three interacting agents: one Buyer Agent and two Store Agents.
  • News Bots, e.g. CNN Custom News - This will deliver the sports score, stock quotes, local weather, and news from CNN and articles from over 100 magazines and news outlets.
  • Newsgroup Bots, e.g. Deja News Power Search - Search for the latest news by filter or search option on all news groups on the Internet.
  • Search Bots, e.g. BullsEye - a unique desktop application that allows users to efficiently find, track and manage information on the Internet just like they do on their desktop.
  • Shopping Bots, e.g. Bargain Finder Agent - An intelligent agent for comparison shopping.
  • Stock Bots, e.g.
    • All Seeing EYE - a javascript-based all-in-one comprehensive business and financial search bot that simultaneously submits searches to over 30 search resources.
    • NavigateOne - An extremely useful site when researching the stock market. Enter the name of a stock and it searches various financial sites, aggregating all the information about that stock located at all the sites.
  • Chatter Bots, e.g. Abby - based on Joseph Weizenbaum's 1966 ELIZA program. At times she can be insightful and at other times a little stupid.

Agent Applications

There are a number of excellent agents that run as stand-alone applications. Some are free and others you have to pay for. Some are offered in a cut-down version for free, or as a fully-featured application for a price (which may not be that much).

Internet agents automate many routine and often time-consuming tasks, giving you more time to spend on other things.

The tasks they handle include:

  • Filtering and downloading Web sites
  • Alerting you when new products are available
  • Alerting you when Web sites have been updated
  • Automatically downloading product updates and bug fixes

Some agents can even let you know when a Web site has been removed from the Internet or when one has been moved to a new location.

Agents may not be the ultimate solution for keeping up with the constant evolution products and Web sites, but they do go a long way towards helping you stay as up-to-date as possible with the latest and best products that the Internet has to offer.

Examples Of Agent Applications

  • Copernic

    Copernic is one of the best multi-site search utilities currently available on the Internet. There is an ad-supported free release and two other versions you pay for. Copernic offers the ability to concurrently search up to 130 search engines and specialized search directories for hard-to-find Web sites, e-mail addresses, and news articles. You don't need to bother with the specific rules of each search engine because Copernic does it for you. It automatically updates its list of search engines.

  • MataHari

    Search nearly 600 sources at once and choose from 60 information channels tailored to your information needs. This easy-to-use desktop search tool lets you use complex or plain-English queries to retrieve target pages from multiple search engines. Mata Hari provides thorough and intuitive results handling.

  • WebFerret

    WebFerret is the World Wide Web application of the NetFerret suite of search clients. This particular ferret will help you find Web sites anywhere on the Internet. Just input one or more keywords, and WebFerret will contact various search engines on the Web to find sites that match the keywords. Search engines contacted by WebFerret include AltaVista, Lycos, Search.com, Infoseek, and WebCrawler as well as some of the lesser know engines like Euroseek, AOL NetFind, LookSmart, and Galaxy. WebFerret is available, an ad-supported, for free, or you can pay for the professional version.

  • NetAttache

    NetAttaché claims to be the Internet's preeminent second-generation Web agent. Does it live up to that claim? The answer is a definitive yes. By integrating unique and powerful features that bring the Web to your desktop with a minimum of effort, NetAttaché is indeed the consummate Web agent. NetAttaché offers Web site archival capabilities, personal spiders that automatically download Web-related information for you, and routine comparison checks for notifying you when your favorite Web sites have been updated. The innovation that distances NetAttaché from the competition is its filtering agent, which will automatically scan Web sites and download only the information important to you. NetAttaché complements its impressive set of features with a good interface and easy to understand instructions.

  • BullsEye

    There's not much more you can say about BullsEye that hasn't already been said about Copernic: BullsEye matches that package feature-for-feature. BullsEye has a really rich set of features and capabilities that make it very useful for someone who needs to track a lot of information on the Web and keep it well organised. If your needs are more modest, one of the other, less fully featured programs might be a better choice, but once you've used BullsEye you don't really need to look elsewhere. Where BullsEye really shines is in the tools it provides to work with search results. You can view documents in your browser, sort your results in a number of different ways, and annotate any result with your own notes. You can also generate nicely formatted reports that can be viewed in your browser, saved to a file, or emailed.

Finally

BullsEye 2

This free intelligent desktop portal gives you instant access to relevant, targeted and personalized information by leveraging over 700 of the best search engines and highly targeted databases on the Web. One reviewer said, 'BullsEye is a researcher's dream. It's the most powerful search tool I've come across.'

Conclusion

Although current agents and bots are good, they could be better. No one package will give you everything that you need, although NetAttache and BullsEye come really close. Future agents will be even more impressive. Features currently under development are things like:

  • Cross-Language Text Retrieval (CLTR) - Your query is converted to another language, searched in that language, retrieved, and presented to you in your own language.
  • Natural Language Processing (NLP). - Queries will be entered as ordinary English phrases.
  • Advanced Knowledge Management - Data mining, filtering, database storage and data sharing will all be integrated. This will include Web sites, documents, email and discussion groups.

 

 

Exercise Number 13: Using Search Software

Suggested time for exercise = 10min SHOW HINTS

In what way are Intelligent Search Agents different from Standard Web-based Search Engines?
Compare the 'discussion group search' from AltaVista with Deja News.

'Stock Bots' can improve your ability to monitor the financial world, and possibly improve the competitiveness of your company.
In what way is this so? Have a look at NavigateOne.

InvisibleWeb.com allow you to search for information not indexable by Search Engines.
Can you find your own email address(es) using their search facilities?

Deja News Power Search allows you to search Newsgroups, Forums and Discussion Groups.
Can you find a mention of BullsEye 2 in any of these? What are people saying about BullsEye 2?
Next section