Tecnica Project Report — Mary Engle

Assisted At:

Instituto de Investigaciones Economicas y Sociales (INIES)
Apartado C-16
Managua, Nicaragua

Background

INIES is an independent research institution founded in 1981 to contribute to viable regional alternatives to the continuing economic crisis in Central America and the Caribbean region. Working closely with 16 other member institutions in the Coordinadora Regional de Investigaciones Économicas y Sociales (CRIES), INIES researchers pull together raw data for analysis, seeking solutions which respond to the needs of the majorities, with emphasis on regional solutions rather than those imported or imposed from outside the region.  They also analyze Nicaraguan internal census data to point to present trends.

The Documentation Center

In 1982, the INIES Documentation Center began its specialized collection of materials relevant to economic, political, and social issues in the region.  With the goal of serving the researchers within the Institute and throughout the Caribbean region as a major reference center, it located within the region the valuable collection of bibliographic information, including books, articles, conference proceedings, theses, clipping files, and newspapers relevant to the topics of interest. Its current collection of 7,000 holdings in closed stacks represents a sizeable databank of macroeconomic information on Nicaragua and the Caribbean.

Amilcar Turcios, Director of the Documentation Center, oversaw this TecNica project for INIES.  An earlier report by Valerie Roach, a volunteer at INIES, described in detail the Documentation Center needs and automation goals.  Entitled, “INIES Automated Documentation Handling System—Global System Requirements,” the report was very useful as background for this project.  It is available in English at INIES.

ISI’s Donation

Before coming to Nicaragua, I wrote a proposal to the Institute for Scientific Information (ISI) in Philadelphia, maker of SciMate. ISI donated two SciMate programs, Personal Data Manager and Disk Utilities for the KayPro 2 to INIES. It should be noted that ISI does not market a production version of SciMate programs for the KayPro 2. They concentrate on KayPro 4 and later models. The copy of SciMate installed at INIES is, thus, a non-production version. The KayPro 2s may not be able to take full advantage of the SciMate capabilities, although I found the program to function normally and completely. ISI also offered to send regular updates to the pertinent SciMate programs. The KayPro 2 version may have some minor incompatibilities with later update versions, but I do not expect this ever to become a serious problem.

Previous Automation Projects

A previous attempt at automation had as its goal the production of printed products—bibliographies, reference publications, listings and catalogs. Online searches were a secondary requirements.

Canadian programmers contributed a database program for data entry. The database used fixed fields on considerable length to accommodate the longest titles, multiple authors, and other bibliographic elements. This design tended to waste enormous amounts of space, limiting storage to 50 records per single-sided diskette. (KayPro floppy disk drives apparently only read single-sided diskettes.)

Although a rewrite of the Canadian program over a year ago increased storage capacity by 100%, much more was needed. The computer project was stopped when the Canadian programmer agreed to prepare a new program that made better use of space. To date, the program has not been complete and installed at INIES.

INIES staff has no printed documentation of the Canadian system, only about 22 diskettes of data.  They ceased inputting data into the old system in July, 1984, shortly before the arrival of Beth Garcia, a North American librarian who is now head of Technical Processing.

Approximately 1100 of INIES’s records currently exist in machine-readable form on Osborne diskettes. INIES has a KayPro program called MFDISK, which allows the KayPro user to read diskettesd produced on a number of other machines, including the Osborne.  The diskettes of machine-readable records are available to both the KayPros and Osbornes in the office.  I was told that INIES records average about 625 characters per document. I find this average to be inflated, considering the records that I saw, which were more like 300-350 characters per record

SciMate Installation

On a KayPro 2 microcomputer, I installed and demonstrated SciMate, a search program and database manager specifically designed to handle bibliographic information.  SciMate uses variable length fields to accommodate data elements of any length with minimal overhead in disk capacity. At this time, SciMate is the most advanced database manager for scientific and academic documents for the personal computer.

One demonstration included only INIES staff, while two others were also attended by representatives of the Banco Nacional, INEC, DNI, CIERA, CIDCA, CREALC in Mexico City, and others.

Toward the end of the first week, I received word from Amilcar Turcios that key staff had approved the installation of SciMate at INIES.

Beth Garcia and I designed formats for the data in templates, which are collections of field names describing a single record. (Each template represents a particular document type.) We developed separate templates for serials, articles within serials, monographs, conference proceedings, and theses.  INIES may wish to add new templates as their collection grows and encompasses other document types, or if they choose to refine existing ones. For example, INIES catalogs analytically, preparing catalog entries for book chapters and other substrata of document groups.

Over the course of the two weeks, I trained Beth to use SciMate features and utilities, iwth the exception of the File Repack facility. File Repack is not likely to be necessary for some time; however, it is an important utility for conservation of diskette space and should be investigated in the future. Beth learned SciMate and the general use of microcomputers very quickly, and I have no doubt about her ability to carry this project through in the future. Beth and I also trained Maritza Huembes, another documentation center staff member, who will likely do most of the data entry of new records into the system.  Beth will continue to work closely with her until she is more comfortable with SciMate and attentive to quality control.

Although SciMate menus and help screens are in English, the template fields and associated codes are all in Spanish, simplifying data input and output for the Nicaraguans. Before coming to Nicaragua, I had the help of Elissa Miller, an Oakland librarian in a Spanish language library, translating many of the SciMate menu screens and in compiling a list of translations of technical terms commonly used by SciMate. I left multiple copies of these materials at INIES.

Beth Garcia and I only succeeded in getting a small sampling of about 30 records into the database while I was there, since we deemed it more useful to enter clean, usable records from the beginning, rather than filling a database with brief demo records. INIES staff will face a period of getting acquainted with SciMate, during which input will be slower, with an eye to quality control. As familiarity increases, so will the input rate. They hope to have a large number of records in the database by July. With such a small sampling of records available, it is difficult to estimate diskette storage capacity precisely, but the SciMate documentation (manual p. 3-119, Table 15) suggests that, given 191K Bytes KayPro disk capacity, a single diskette will hold at least 300 records at about 600 chars. per record, and at least 400-500 if my earlier estimate of approx. 300 chars./record is correct.

Producing Bibliographies

SciMate outputs data in limited, predefined report formats. The appearance of the columnar reports is suitable for certain of INIES’s needs, such as informing documentation center staff and researchers of recently acquired materials. There was a great need for an output formatting program to produce high-quality bibliographies. Another TecNica Volunteer, Ken Fishkin, wrote a BASIC program to print bibliographies in three formats in use at INIES.

To produce bibliographies, the user first creates a text file of citations using SciMate, passing it as input to Ken’s program, called BIBLIO. BIBLIO prompts the user for the name of the input file and choice of output format, all in Spanish. Documentation in the source code is in both English and Spanish. The program is quite an enhancement to the outputting capabilities, and Ken’s work, under a great deal of time pressure, was a considerable contribution to this project.

Possible Future Projects

The single-most important project at this point is the conversion of the 11090 records from the old automated system to the SciMate database.  We were not able to accomplish this before I left, but I may be able to get help from programmers in Berkeley to prepare a program to read the Osborne/KayPro data diskettes and write a file usable by SciMate to automatically load the records into its database. Ken Fishkin has agreed to consult on this.

I had discussions with another INIES volunteer, Gonzalo Zepeda, of the National University of Mexico, about preparing a parallel database of brief key records linked to their full record counterparts by an accession number. Key records provide faster access to holdings in queries where full-text searches are not required. This would be useful as a raeference tool, both in know-item searches and descriptor “browsing,” and will become increasingly appealing as INIES’s SciMate files become numerous and lengthy (abstracts may be added in the future). This possibility will be explored more thoroughly. Sr. Turcios expressed interest in it as well.

I have ideas on how SciMate itself would be able to produce the abbreviated key reocrds as a function of its ability to load a subset of an existing database into a newly created database. I am developing these ideas with ISI representatives now, and should have it worked through in the near future.

It would be useful for the BASIC programmer to check in a few months time to see if the BIBLIO pgoram is meeting their needs, or if they have encountered any problems with it. We had woefully little time for testing and enhancements.