Previous Table of Contents Next


Search Engines

In addition to the above four components, search engines are constantly being created that help users find the Web sites that store desirable information. WAIS (http://www.wais.com), InfoSeek (URL>http://www.infoseek.com), Yahoo (http://www.yahoo.com), WebCrawler (http://www.webcrawler.com), Lycos (http://lycos.cs.cmu.edu) and SavvySearch (http://guaraldi.cs.colostate.edu:2000) are often used for Web searches. These search engines organize their own databases, start their own search mechanisms to support queries ranging from simple query statements to complex formations and even natural-language queries, and they return a list of URLs. Without these searching machines, finding a list of desirable URLs from the vast, unstructured, uncoordinated Web resources is time-consuming and could take the users months of point-and-click navigation to assemble.

WEB LIMITATIONS: UNANSWERED BUSINESS QUESTIONS

With its capabilities, the Web has been able to facilitate electronic business transactions. Product promotion, customer support, and electronic publishing are a few examples of functions in which Web technology has been successful. Nevertheless, from a business perspective, four fundamental questions, described below, remain unanswered. These questions have prevented many corporations from carrying out business on the Web.

  Is the Web navigation mechanism effective? The Web employs the hypertext mechanism for navigation. For a typical query, the user is often required to do some clicking on the mouse to reach a desirable Web site. Once arrived at a Web site, more clicking is required before information may be obtained. Although searching engines have alleviated some of the difficulty in reaching Web sites, users are still required to do a lot of clicking and bouncing around from page to page following pre-designed links. Such a simple navigation mechanism is not flexible enough to give users more specific information and quicker responses to business queries.
  Is the Web data structure adequate to support information reporting and query responses? The Web employs a hypermedia data structure in which information is stored in small chunks, called pages. Text documents and other object-oriented data are fitted into these pages. However, traditional record-based business data and numerical data are not suitable for storage in pages, partially because business data, if stored in pages, cannot be easily accessible on a record-by-record basis. In addition, HTML is just not powerful enough to handle record-oriented business data, nor does it allow user-controlled queries to be easily formulated. Consequently, key information cannot be provided under present Web-based data structure.
  Can enterprise data or legacy data be available on the Web? To date, enterprise data — mostly transaction oriented — is stored mainly in mainframe computers. Security and performance concerns are two major reasons why enterprise data is mostly inaccessible from the Web. Methods and techniques are being developed to bring mainframe-based data into the Web. At present, these methods and techniques are not feasible and therefore transaction-related information on order status, invoice, bill of lading, and payment will mostly remain unanswered.
  Is the Web suitable for mission-critical business applications? The Web is not set up for on-line transaction processing and has failed to meet the standards of security, performance, backup, and user management. For example, Web technology is inadequate to perform the five security-related tasks (i.e., authentication, certification, confirmation, nonrepudiation, and encryption); therefore, an interactive transaction between trading partners is not reliable. Besides the security concern, other key factors have also contributed to a lack of Web-based mission-critical business applications. These factors include stateless conditions during transaction processing, questionable bandwidth to handle real interactive transactions, and lack of user preparedness for electronic commerce.

INTRANETS

Despite a lack of legacy data on the Web and immature Web technology for effective transaction processing, an increasing number of corporations are now turning to the Web as their IS solution for addressing business problems within corporations. It is predicted that internal Web or intranet usage will surpass external Internet usage by the year 2000. The key factors for adopting intranets are open platform standards (e.g., HTTP and HTML), ease of installing Web servers and using Web clients, and multimedia capabilities.

The range of intranet applications that can be developed is virtually unlimited. Currently, corporations are deploying intranets as a way to organize their internal communications. Examples of these intranets are:

  Web-based internal E-mail systems.
  Project planning, monitoring, and reporting.
  Forums for brainstorming, collaborations, or problem solving.
  Delivering marketing materials, training materials, or software products.
  On-line customer information queries.
  On-line human resource queries on employee benefits, company policies, personnel information, employee and telephone directories, job listings, and training and education workshops.

One main concern of deploying intranet applications on the Web is security. Currently, several measures are being installed, including firewalls. Most firewall products focus on keeping external Internet users from getting into intranet applications. Others ensure that users are authorized to access the information they seek.

INTEGRATING THE WEB AND ENTERPRISE SYSTEMS

The process of integrating the Web and enterprisewide systems or building some intranet applications can be approached from two directions. One involves converting enterprise data into hypermedia pages. The other involves building a link between these two systems. Regardless of which approach is used, the goal remains the same; that is, making enterprise data and the various business applications accessible through Web browsers. The use of Web browsers eliminates concerns about heterogeneous hardware and various operating systems over the Internet and intranets as well.

Building links to tie the Web and enterprise systems into a coherent system is much more feasible than converting to hypermedia pages. This is partially because the linkage programs will not interfere with the normal operations of enterprise systems for supporting day-to-day business activities and management decisions. Both researchers and vendors have been placing their emphasis on developing architectures and tools to support the construction of the linkage programs.

Converting to Hypermedia Pages

Enterprise systems are characterized by a variety of data structures, including traditional flat files, relational databases, IMS databases, object-oriented databases, and special package-related files (e.g., spreadsheet files, song clips, and photo images). Theoretically, this data can all be converted into hypermedia pages to support applications ranging from information inquiry to transaction processing over the Web.

Although current technology is not mature enough to support certain tasks effectively over the Web (e.g., complex interactive transaction processing), migrating key enterprise data to the Web will certainly give customers speedy query responses for such applications as marketing and electronic cataloging.


Previous Table of Contents Next

Copyright © CRC Press LLC