This article studies the history of Web portals widely used in business-to-business and business-to-consumer Web applications in the late 1990s. Web portals originated from Web search engines in the early 1990s and evolved through Web push technology in mid-1990s to its mature model in the late 1990s. This article also compares Web portals with other popular media, such as radios and televisions, for their audience base and content broadness. As of January 2003, only a few libraries had adopted Web portal technology despite the widespread use of my.yahoo.com-type Web portals in the business sector. The article examines several reasons for the lack of portal development in libraries and concludes with a set of Web portal development guidelines for academic libraries. Some of the pioneer library portals are also discussed, as well as the California State Government, the first government portal to offer customization and financial transactions for individuals and business. This article concludes by probing a more fundamental question about general information storage and retrieval processes. In the last several hundred years, libraries primarily built hierarchical data structures and librarians provided information service without any search engines. In the past ten years, Web business communities have primarily worked on developing fast search engines for information retrieval without paying much attention to data structure. Now with the exponential growth of data on the Web, it is time that librarians and computer engineers work together to improve both search mechanisms and data structures for a more effective and efficient information service.
What is a portal? "Portal" has been the buzzword of the networked age since 1997. Portals were so popular in business-to-business (B2B) and business-to-consumers (B2C) applications that the business world borrowed an old jingle: "I'm a portal, he's a portal, she's a portal, we're a portal, wouldn't you like to be a portal, too?"
Portal derives from the medieval Latin word portale, meaning "city gate." American Heritage Dictionary defines a portal as "a doorway or an entrance, or a gate, especially one that is large and imposing." New definitions for portals in the networked environment can be found on many Web sites. A synthesis of these new definitions is as follows: a Web portal is a doorway that can be customized by individual users to automatically filter information from the Web. It typically offers a search engine and links to useful pages, such as news, weather, travel, and stock quotes. A portal can also be defined as a customizable Web search engine to reflect the MY trend in current Web development. The platform for a portal Web site is a search engine, but a portal is different from a general search engine in that it can be customized by individuals for automatic, constant search for specific information, and it can deliver the results to individuals in a predefined way. A customizable search engine is unique to the user; it is different from anyone else's.
The very early history of portals used by librarians can be traced back to the 1960s, when the first digital version of Index Medicus was created. (1) Some science librarians may still remember the customized weekly search in Medline for medical researchers and in INSPEC for physicists. This kind of canned search was predefined offline first by scientists and librarians together with a set of criteria. The canned search was performed by librarians against the weekly updated database tapes on IBM mainframes. Finally, the search result was delivered to scientists for the most recent developments in related fields. In the business community, CEOs often had various Executive Information Systems (EIS) before the Web came into existence in 1992. EIS was developed to provide top decision makers with broad, diverse content according to previously defined criteria. Both librarians' canned searches and the EIS service can be seen as human-controlled portals as they provided customized information in a timely manner through human mediation.
The history of current Web portals can be traced back to the Boolean search technology developed in 1994 and 1995. (2) Due to the exponential growth of Web pages, users cannot locate Web files by conventional means, such as using directories like those in phone books. Search engines offer document content with full-text indexes and direct links to the documents in the Web environment. Initially, most Web users were researchers and educators with high-speed Internet access. They were very excited about the advent of the search engine. Academic librarians all over the country started to offer college-level library instruction courses on Web search engines as early as 1992. When more and more people started to use the Web, not just for research but for entertainment and daily information gathering, it became evident that giving people a complicated search command language to find popular information was not efficient. In order to address the average user's frustration and reduce the seek time for relevant information, the search sites added the function of channeling or categorization--filtering popular sites and documents into preconfigured groups. This categorization is a step further from the general Web site; the Web provides broad information for a broad audience and categorization provides narrowed information for a broad audience. The concept of …