Transforming Print Encyclopedias into Successful Electronic Reference ProductsSkip other details (including permanent urls, DOI, citation information)
This work is protected by copyright and may be linked to without seeking permission. Permission must be received for subsequent distribution in print or electronically. Please contact firstname.lastname@example.org for more information. :
For more information, read Michigan Publishing's access and usage policy.
Content and Added Value
The range of content to be included might not be as obvious as you might expect. Should you include prefaces, various content lists, indexes? Editors can advise here, as well as on the details of the larger puzzle: What can we add to this new electronic database to make it more useful? As I mentioned above, one way we have made our books more useful in electronic form is to combine them with others on like subjects. But a large database can be unwieldy. How will the user approach the content? Will every search give 300+ results, and if so, what is the user supposed to do next?
The development team will have to do some research to find out which searches or ways of using the product are most likely to be the most popular, and the editor can edit search results to meet users' needs. We have added special searches and lists of report topics to our products to make them friendly to high-school students, and we have added hundreds of pictures to works that were not illustrated at all in print. Picture research was supervised by the editor, who learned to choose pictures that not only elucidate the text but also look clear at computer-screen resolution.
"Don't plan a lot of expensive features without first checking to see whether the market really wants them"
In the Scribner Writers CD, we added headnotes that serve as abstracts to long bio-critical articles, so that the user does not have to read a long essay to find out whether she wants to continue investigating in a particular direction or not. For example, if you are reading the article on Ralph Waldo Emerson, and Thomas Carlyle's name comes up, the word "Carlyle" is a link to the article by that name. Click on it, and you do not go to the beginning of a 12,000-word article, where you might have to read a long passage to figure out who Carlyle was; instead, you go to a headnote that gives you his birth and death dates, his nationality, ("Aha," you say, "He was not a New England Transcendentalist like Emerson; he was British"), a phrase or two about what he was famous for, and his picture. If you would like to continue reading about him after glancing quickly at that information, you can. Or you can go back to Emerson and continue reading about him, without the flow of the Emerson article having been drastically interrupted. The headnotes serve as a bridge from one article to another and add a new layer of context to the articles. The in-house project editor who originated that idea hired freelance writers, researchers, and copy editors to create the headnotes. Every special feature we have invented has taken much editorial thought and time to devise and execute.
The question of how much can be done automatically through programming ("programmatically") comes up. Should every article title embedded in text be automatically hyperlinked to the article itself? What if the article is about the person "Napoleon" and the text refers to the dessert? What if the first two names of "George Washington Carver" are linked automatically to an article on George Washington? What if automatic links are so tangential that no editor or indexer would ever choose to waste a reader's time with them? Then our product is not making sense, that's what. If automatic linking is to be done and then revised, time must be allowed for testing and fixing. If we forgo automatic linking altogether, editors or indexers must identify each useful link by hand. If we are dealing with many millions of words of text, that might be prohibitively costly or time consuming.
To prepare usefully limited search categories, the editor will probably find himself involved in the creation of a set of descriptors about each reference article called "metadata." Articles will be put into categories based on predictions about users' interests determined by surveys and other market research; each article must be considered individually and placed in one or more of those categories. Some companies, especially aggregators, find it useful to use Library of Congress categories; at Scribners we use simpler, curriculum-related ones. The more those decisions can be considered and reconsidered, the more rich and useful the future electronic product becomes. For example, an indexer puts all the Dictionary of American History articles on court cases in the "Law" category; the editor spots Brown v. Board of Education and adds it also to the "Civil Rights" category. "Law" and "Civil Rights" are now part of the metadata for Brown v Board of Education. Metadata also includes information about the print origin of the product (title, volume number, pages, contributor, number of illustrations, etc.), some of which might actually appear on a screen somewhere in the product and some of which you are saving for use in the creation of other electronic products or print spin-offs in the future. If the metadata can be organized hierarchically, it can be used to create broad or narrow searches.
Again, the importance of editorial involvement and oversight in creating special features cannot be overemphasized, nor can the importance of keeping in touch with your market at all times. For example, don't plan a lot of expensive features without first checking to see whether the market really wants them; a sound track, for example, is probably unnecessary for a product to be used primarily in libraries where quiet is usually the librarian's aim.