The annual conference of the Society for Scholarly Publishing (SSP), held in Washington, D.C., at the end of May, included papers on all aspects of electronic publishing--from the preparation of machine—readable manuscript by authors, through electronic-assisted publishing and printing, to the production of journals in electronic form, and electronic document delivery.
ELECTRONIC MANUSCRIPT PROJECT
In a paper presented at the conference, Aspen Systems described the work it has undertaken to develop guidelines and standards for the electronic preparation of manuscripts by authors. This work is being undertaken on behalf of the Association of American Publishers. The Council on Library Resources, Inc. is a major contributor to the project.
The Electronic Manuscript Project grew out of the realization that savings could be achieved if publishers could accept and use the electronic versions of manuscripts that are prepared by an increasing number of authors. (A survey by University Microfilms indicates that 50 percent of the dissertations produced today are prepared in machine-readable form.) The development of standards in this area would diminish the need for rekeying and make it simpler and less expensive to utilize translation programs that take text from word processors and format it for use in publishers' electronic systems.
The project aims to identify the component parts of a document and develop appropriate generalized tagging/ generic coding practices to delineate these elements. The results are expected to assist authors, publishers, editors, and printers, and also to encompass the needs of other groups such as librarians. Graphics experts are also being consulted for the definition of other document elements such as tables.
[A requirements analysis as well as related industry surveys have been completed.] Publishers report that 30 percent of their authors now prepare manuscripts in machine-readable form, and they expect that by 1985 some 50 percent of manuscripts would be available in this form. Of the available machine-readable manuscripts, publishers estimate that the machine-readable characteristics of only 40 percent are currently being used in the publishing process.
The results of an authors survey differed from those of the publishers in relation to the estimates of the availability of manuscripts in machine-readable form. Authors surveyed said that .60 percent were now producing their materias in this way. Forty percent of those authors reported having been asked to follow some sort of manuscript preparation guidelines. In this area, the authors want consistency. The problem is that different publishers have different manuscript preparation requirements.
A 15-page summary of the survey results is available free from:
Electronic Manuscript Project
Association of American Publishers
2005 Massachusetts Ave., N.W.
Washington, DC 20036
Aspen is now identifying the elements within manuscripts. A random selection from a list of several hundred elements designated as requiring tagging by various groups involved in the project includes: author, title, conference/meeting, imprint, physical description, notes, unique numeric identifier, full text, formula, table, footnote, caption, list. A draft of the elements and the generic tagging to be used to distinguish each will be available this summer. Sites for the testing of the guidelines are being identified.
The range of groups involved in consideration of the guidelines is wide--archivists, data base producers, and members of the library and abstracting and indexing communities have all been included.
Librarians have focused on the requirements for monographs and serials, there being a consensus that maps, music, and AV materials are unlikely to be submitted in machine-readable form. Patents are being dealt with by another group. In addition to identifying elements that have been defined because of their need to be used, library advisers have asked for the identification of some elements so that material in these categories-advertisements and author's acknowledgments, for example--can be avoided. In addition to enumerating document elements that should be identified, suggestions have also been developed for different levels of tagging--analytics,monographic level, collective level—and the identification of unique and repeatable elements. In some data elements the level of tagging sought is great—the component parts of an author's name for instance. Librarians have emphasized the need for clear linkages between different levels of document elements. The abstracting and indexing community had a strong interest in the inclusion of addresses for obtaining reprints.
BOS--BOOK ORDERING SYSTEM
The main focus of the presentation by Sandra K. Paul, President of SKP Associates and Editor of EPB, Electronic Publishing and Book Selling, was to discuss the microcomputer-based options that allow small publishers to tie into book distribution systems. Particular attention was paid to the features of BOS--the Book Ordering System developed by the American Booksellers Association. Other systems mentioned were BAS--the Bowker Acquisition System—which provides online access to the machine-readable version of the Books in Print file and supports the generation of orders in either paper or machine—readable form. A similar system is offered through the Canadian Textbook Agency, and the National Association of College Stores is setting up a system called NACSNET to handle the ordering of new and used textbooks and other college store items, such as Tshirts, coffee mugs, etc.
BOS has been designed to provide bookstores with an inexpensive way of speedily relaying orders to publishers. A microcomputer is used to prepare the orders of f line. These may be printed and mailed or electronically transmitted to the BOS mainframe in New York from whence they are distributed to the designated publisher or wholesaler. The system has been operating in test mode since mid—February and was released for implementation in bookstores on July 1. Initial plans call for the signing of 100 stores a month for the first 12 months.
An attractive feature of the system for both bookstores and publishers is that orders are billed to BOS and BOS pays the publishers directly. BOS then collects the money owed from the bookstores, relieving the publishers of payment delays and collection hassles. Items are shipped to the individual bookstores by the publishers. The bookstores benefit in that this approach gives the individual stores the combined purchasing power to take advantage of quantity purchase discounts from publishers.
While the BOS mainframe houses a file of data on the items published by participating publishers and the system requires that publishers update this information monthly, these data are used only for mainframe processing and transmission of the orders. As presently conceived, BOS does not provide participating bookstores with access to an online data base that can be queried to determine which publishers provide which titles. However, there are a number of ways in which the system hardware can be used to provide this information. A machine-readable version of Books in Print is available on Dialog and the hardware can be used to access this file. Such access, however, would incur data base access and telecommunications charges. A number of publishers are developing machine-readable versions of their publication lists for distribution for local use on the BOS system. Houghton Mifflin is distributing a floppy disk of its titles to those stores that are using BOS. A possible future development of the BOS system might provide direct online access to publisher's list data.
Another approach is being promised by the Library Corporation, which is the joint development--with Library Systems and Services, Inc.--of the laser videodisk MARC publication described in the previous issue of LSN. The company is promoting the Any-Book/MARC disk as a standard peripheral to the BOS system.
BOS uses SANs (Standard Address Numbers) to identify publishers and bookstores. Orders are generated in the BISAC format. Electronic invoicing is also supported.
The system is designed for an IBM PC-compatible micro. Bookstores using the system will be able to purchase the micro an,da printer for about $1,149, or to lease the hardware for $80 a month. The monthly software charge is similar to that for hardware.
For publishers, there is no charge for participating in BOS. Funded by the American Booksellers Association, BOS will be offered to any organization that is a full member of the association. No libraries are currently full members of the association. (If such membership should be possible under the rules of the Association, and libraries chose to avail themselves of it, then they could sign up to use the system.)
It is intended that the facilities of the system be made available to book wholesalers as well as publishers and bookstores. Negotiations are under way with Baker & Taylor, and Ingram.
[Contact: BOS, 160 Fifth Ave., New York, NY 10010; (212) 206—8212.]
IAC/OCLC ELECTRONIC DOCUMENT DELIVERY EXPERIMENT
In another SSP Conference presentation, Information Access Company described its primary periodical full text data base that it has developed under license from 47 trade publishers. OCLC and lAC have been working together to develop an overnight electronic document delivery system using this data base, which is presently available on Dialog under the name “ASAP.” It is intended to use the OCLC telecommunications system to distribute article text in response to queries logged on the OCLC interlibrary loan system, providing local printing through printers attached to OCLC terminals. Royalties would be paid to the publishers as is already being done for the ASAP data base.
The IAC full text data bases are keyed conversions of the text of individual articles into machine-readable form. The file consists of machine-readable coded data, not digitally encoded image representations. The full content of each article is reproduced as much as can be accomplished within the normal character set limits of generally available terminals. Normal headings and highlighted text are included, as are references and footnotes, but complete tabular and graphic displays are not presently possible in this format.
The present coverage was designed to provide a “critical mass” of text articles available online, but it is considered only a beginning in full text delivery. The 130+ publications covered in the initial list represent about 30,000 published articles annually, but that number totals less than 5 percent of the articles that lAC indexes annually. The text data bases represent about 250 million characters of text annually——almost the size of the entire set of IAC index data bases. (Statistical analyses undertaken in the 1970s suggested that an effective index is about 1/30 the size of the material it indexes; IAC seems to be confirming that in 1984.)
IAG believes that for electronic article distribution to be feasible, it must combine organized index access with electronic delivery. It is felt that full text information retrieval or searching is both expensive and ineffective.
IAC expects its full text data bases to be used as a searching tool primarily for locating unusual words not found in either the indexing or the title; for casual mention of corporate or organization names; and for other specialized purposes. The 30:1 ratio between the size of the index data base and the size of a comparable text data base means that the relative expense of searching directly in the full text file will exceed that of searching in the index files. Also, because of the wide-ranging nature of the IAC text data bases, the variation in meaning of common English-language words will result in a high level of noise in full text retrievals.
The IAC service is based on the supposition that the combination of index access (which is online in the case of Dialog but may be offline in other situations) and electronic document delivery represents the most powerful and cost-effective way to provide electronic distribution of the type of information IAC covers. In 1983 OCLC and IAC entered into an agreement to explore the feasibility of distributing the IAC full text data bases through OCLC. The concept is to use the OCLC interlibrary loan subsystem to trap requests for articles that are available in the IAG electronic full text data base. IAC would become, in effect, an electronic library on OCLC, responding to any library's request for a document. What would be unique about this system is that the requesting library would know that this “library” had the article desired in stock, and would respond overnight. The actual text delivery would take place through the OCLC telecommunications system, and would be available the next morning at the OCLC terminal. Because the system utilizes the OCLC and IAC computer systems and telecommunications networks at night, when online use is at a minimum, it is hoped that the marginal cost will be low enough to allow an attractive per-article delivery price considerably less than that of instantaneous online delivery.
This project is only one of a number of methods of document delivery with which OCLC is involved. The others (which involve nonelectronic delivery) include:
- Through University Microfilms, Inc., electronic requesting and mail delivery of articles from a wide variety of publications, including scholarly and trade materials.
- Through the Universal Serials and Book Exchange, Inc. (USBE), electronic request and mail delivery of individual back issues and runs of 7,000 active periodicals.
- Through Information on Demand, Inc. (on a pilot basis), custom photocopy with mail or express delivery of essentially any periodical in many research libraries.
