Add to bookbag
Author: Ludmila V. Emelianova
Title: An Information Retrieval System Mutifunctional Interface: An Instrument and Means to Catalogue Handwritten Stockpiles
Publication Info: Ann Arbor, MI: MPublishing, University of Michigan Library
August 2001
Availability:

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 mpub-help@umich.edu for more information.

Source: An Information Retrieval System Mutifunctional Interface: An Instrument and Means to Catalogue Handwritten Stockpiles
Ludmila V. Emelianova


vol. 4, no. 2, August 2001
Article Type: Article
URL: http://hdl.handle.net/2027/spo.3310410.0004.206
PDF: Download full PDF [17kb ]

An Information Retrieval System Mutifunctional Interface: An Instrument and Means to Catalogue Handwritten Stockpiles

Ludmila V. Emelianova

An important component of information retrieval system (IRS) is the form of information delivery implemented in interface software package. Establishing a multipurpose IRS interface for research cataloguing of manuscript collections is a primary concern because successful implementation in this case would largely determine viability of the whole system.

The IRS incorporates a set of functional options used in research cataloguing of manuscript collections. Among them are a work lists for manuscript record input and updating, forms for relevant directories, retrieval query lists, a selected information display, screen forms for imaging specific document characteristics, an interface facilitating index file building and viewing, forms to display information for different user categories, interface for occasional printing by user request.

.01. INTRODUCTION

An important element of information retrieval system is a user interface incorporating the total data display and a set of functions and services. Even the most highly functional software in terms of electronic catalogue retrieval speed, as well as the system development efforts may prove useless with a non-user friendly interface. This is why programming experts place special emphasis both on the design, and on the multiple functionality of the system involved.

The problems we propose to solve with computing inevitably involve some difficult elements. On the one hand, the programmer should conceptualize the problem and continuously check his concepts against the user's response. When this practice is ignored in development, the programmer's lack of knowledge in the sphere of application will affect the final development stage, resulting in additional labor costs for finishing. On the other hand, if a user is involved in the program development stage, he or she will also begin to understand the system facilities and formulate his or her demands with better awareness. Not infrequently, the user would adjust his or her demands in the course of development because, as the problems to be solved are often modified as the project is implemented.

.02. GENERAL INFORMATION ABOUT "DEPOSITORY 1.0"

The information retrieval system described here, "Depository 1.0" is designed for an electronic catalogue of research records of archival documents. The concept underlying the program's functioning provides for the database serving as both a means to record the manuscript collections, and an effective research instrument for manuscript material to produce new and comprehensive knowledge of a document. The development of the system took almost five years to complete. The long development period was governed by the need for the best possible version to meet user demands on the one hand, and standard requirements on the other. Special attention in developing the interface was paid to its user-friendly quality, integrating into the system those utilities likely to be familiar to the user, such as word processors to handle printed lists and multimedia programs for digitized data. For each data representation, we intended an intuitively understandable interface.

The "Depository 1.0" program includes the following functions:

  • browsing of electronic catalogue records both in conventional form and distributed in selected fields;
  • input and adjustment of manuscript record data and analytical elements in a UNIMARC-based format;
  • reference building as the electronic catalogue is being implemented;
  • browsing using graphic packages of digitized image files linked to relevant record data elements;
  • searching and retrieval over the input work sheet fields;
  • building printed lists and indexes.

The information retrieval system is based on the UNIMARC format (<www.ifla.org/VI/3/p1996-1/sec-uni.htm>) whose Russian version for printed publications is RUSMARC (<www.nlr.ru/rba/rusmarc>). All fields and sub-fields starting with 9 in this format are used by our Library. The fields are intended to, without disturbing the overall format structure, represent the elements of the document to be recorded for use in a specific organization. We have made great efforts in structuring and field selecting for all relevant information to provide research-based description of archival material in machine-readable form. In doing so, we were guided by the following considerations:

  • field structure providing maximum targeted searching;
  • input data represented in conventional research form as research record;
  • correct retrieval in research;
  • building required printed forms;
  • potential addition of digitized data.

.03. DETAILED INFORMATION ABOUT "DEPOSITORY 1.0"

The Main Box

The "Depository 1.0" main box is divided in two parts. The bottom part shows the records in the alphabetical order with sorting both by call number and initial record line. The upper part contains full-text description of the document in to the current record. This information arrangement will assist the user in fast orientation in the electronic catalogue. A provision is made for viewing the document description in several forms, both conventional-research and in selected format fields.

The Input Work Sheet

The input work sheet consists of a list of format fields accompanied with prompt help containing information on assigned field or sub-field. To enter recurrent information, for example, in the case of more than one author, the user may select any number of fields or sub-fields. Thus the researcher is not restricted by software facilities in entering required information.

Of course, a certain difficulty exists for the researcher in entering data in format fields. First, fields names are designated by a numeric code, which is unusual for the user. But in practice the designations are easily learned and present no problems. Second, the fractal nature of input information somewhat impedes processing of a manuscript document. But in our case this is compensated by the possibility of formulating complex SQL-queries to provide retrieval terms in research. In addition, information subject to unification is provided in respective directories, facilitating input of recurrent data. The same directories can be also used for targeted electronic catalogue searching relevant indexing for printed catalogues. The mechanism of maximum provision of such directories is also useful to prevent incidental input errors, eventually influencing correct electronic record retrieval.

Searching — Terms

The form for search terms to complete contains the same fields as in the input work sheet. This approach takes maximum advantage of the system's retrieval facilities. Searching is also provided in non-search fields such as notes, for example. But the most successful searching is certainly provided by the reference apparatus. For a researcher entering data in the electronic catalogue, the input work sheet is convenient. No difficulties arise in formulating an SQL-query. Yet a researcher not familiar with electronic catalogue production will face an alternative: either to learn the system retrieval mechanisms, or to fall back upon simplified searching with simple search term formulation and difficult multi-aspect database retrieval.

Searching — Graphics

The information retrieval system does provide for graphic search. This is accomplished by adding the fields to formulate a link to the graphic file, providing for record retrieval. It should be noted that the same mechanism can be also used for sound files. Graphic files are views with a special program listed in the relocation file. In this case the user can choose that graphic program he or she finds most convenient for viewing.

Data Print-Out

Data print-out involves exporting data to the file and viewing in a word processor for subsequent final editing and printing. Thus the proposed system can provide various lists of records for manuscript collections, to be used by the researcher in investigations, and name, geographical etc. indexing for printed catalogues.

Required Duplication

In representing manuscript descriptions in conventional research form, we had to solve the problem of required duplication. This was due to the complexity of automatic formulation of linguistically correct Russian text in screen representation of research-based manuscript descriptions. We therefore had to add in the relevant fields a sub-field for the researcher to edit the automatically built text in accordance with Russian linguistic rules.

.04. CONCLUSION

The interfaces described above, by research-based description, various data display options, and required search modes, make the information retrieval system "Depository 1.0" a convenient instrument for research cataloguing of manuscript collections.

Ludmila V. Emelianova