ProQuest Information and Learning has launched PQNext, a major revision interface that introduces a range of new features to increase precision searching. The PQNext interface supports browsable lists, grouping of results, formatted bibliographies, improved displays, and refined help screens. It also works with OpenURL for authorized linking to full-text sources and lets library licensees co-brand.
ProQuest targets the educational institution and public library markets and hopes the new interface will supply sophisticated searching to satisfy both novice and expert searchers. Developing the new interface, however, is only the beginning as ProQuest plans to restructure all of its technical platforms for future advanced developments.
PQNext offers experienced searchers and librarians a world-class interface for dealing with traditional published literature by exploiting all the value-added components of traditional library products. As Vince Price, ProQuest's vice president of marketing, pointed out, the tool even serves to educate novices about using better searching techniques. However, in a Google world, can this approach suffice? Richard Wiggins of Michigan State University said: "Librarians are put in the unfortunate position of telling people to eat their spinach, that fast food searching isn't enough. But if a vendor could deliver quality material through Google interfaces, they would have an advantage."
PQNext's design responds to the expressed needs of ProQuest's customers. According to Price, the developers conducted careful analyses of usage logs; interviewed hundreds of users of all types, including K-12 librarians, university faculty, and working librarians; and held dozens of focus groups. When previewed at the ALA meeting held in Toronto last June, the audience applauded the Auto Citation Style option, with which users can mark articles online then generate a formatted bibliography in a range of citation styles. PQNext supports a number of citation standards, including AMA, APA, Chicago Manual of Style, MLA, and Turabian.
PQNext also provides article or hit counts, views of indexing or controlled vocabulary through a "Browse Topics" link, and grouping of articles by publication type, such as newspapers, scholarly journals, trade publications, etc. Business searchers can apply key classification and NAICS codes to identify industries, markets, geographic areas, topics, or article types.
Coding and thesaurus terms vary from file to file, making cross-file searching somewhat perilous when the use of a code invisibly limits searching to one file only. Price explained that as part of PQNext's development, ProQuest made some default assumptions. "We have tried to have the system judge what the user is trying to do and make the interface sympathetic to that, but the user can always override it," he said. "The behind-the-screen customization extends to institutional access. For example, though ProQuest uses one giant merged thesaurus, each user will only see the thesaurus terms that apply to the databases selected. It's all tied to licensing. As a user searches the subset of their own subscriptions, features will turn on and off."
Unlike most relevancy feedback functions in other search engines where preset algorithms rule, PQNext's sophisticated "More Like This" option lets searchers designate the specific similarities that interest them. The "Find Similar Articles" screen allows users to specify subject categories, classification codes, author name, journal title, article type, and geographic location. A click on the subject categories will list the subject headings attached to an article. With another click, the user can retrieve items tagged with the subject heading.
The system also offers traditional relevancy ranking and lets users sort by regular or reverse chronological order. After marking items, users can choose to print citations, e-mail articles, or export references to standard bibliographic database management systems such as ProCite, EndNote, Reference Manager, and RefWorks.
A Research Summary feature provides a reviewable list of marked articles, search series, publications visited, etc. Users can edit the summary, add headings and notes, print it, and download or e-mail it as an HTML file. They can even post Research Summaries to Web sites for clients and patrons to use.
OpenURL technology integrates access to the full text of articles from ProQuest collections or those of its partners. It also integrates with e-journals, catalogs, linking resolvers, and document delivery suppliers. Authentication issues—such as offering library licenses to use data sources—are handled through a cascading authentication system each time a link is used. Library branding identifies a database as a paid resource of the library or educational institution that's hosting the searcher and allows users to return quickly to the library home page, if necessary.
All ProQuest-branded products will offer the new interface. In time—and after adjusting for the differences in database structures—ProQuest Information and Learning hopes to bring its other product lines (such as Chadwyck-Healey) into conformity with PQNext.
PQNext has a network of Help options as well as online instructional guides. ProQuest has begun scheduling training programs for librarians that will teach them how to train their own end users. Before using the instructional material, experienced searchers might want to review it. For example, I noticed an FAQ describing Boolean searching operators. ProQuest requires searchers to use an atypical AND NOT for the NOT operator and does not warn them about the dangers of using NOT in full-text searching. Although phrase-making uses the Net style of enclosure by quotation marks, in another oddity, the technique works for three or more words.
Basic Search, the approach most end users will probably try, follows more of a Google-style system but still requires the user to select a database. In addition, it suggests options for type of journal, date range, and more search options, such as specifying publication title, author, what part of the record to search, etc. Price admitted that PQNext was designed in part to teach research techniques "in support of the mission of academic libraries and even the needs of public library patrons."
But in a Google world where users believe they already know how to search and where software houses offer sophisticated post-Boolean search engines capable of relieving users of the burden of learning sophisticated search techniques, is PQNext enough? Apparently not, and ProQuest knows it. Price said: "PQNext is not an end, it is a beginning. We already have a ‘skunkworks' group set up to take us to the next generation. We had to offer these visible improvements to the customer, but internally, we are laying the groundwork for the next generation with evolved features and products. We have begun reconfiguring our vaults, applying content linguistic analysis, building platforms." In summer 2004, possibly even by December 2003, Price expects users will see "quantum change" developments from ProQuest.