Retrospective conversion of library systems

The information storage and retrieval problem has become progressively more serious in recent years, especially in the areas of technology and science, where the volume of data and information is increasing at an unprecedented, nearly exponential rate. Keeping in view the tremendous flow of information, to organize information and to disseminate it in a systematic way, only computerization is the answer. A computer-aided system is more convenient, more flexible, and more comprehensive, and in the long run, more economical. To establish a computer-based system for any library, the first step is to create a database of the existing resources available in that library.

It is necessary to convert the bibliographical information of the manual card catalogue to machine-readable format for use with an automated library system. This process is well known as Retrospective Conversion (RECON). Retrospective Conversion involves a considerable amount of work, essentially acquiring existing databases and keyboarding all the existing entries of records, error checking, etc., or to get the job done from commercial vendors and creating a bibliographic database and transferring the records to machine-readable form. RECON is a costly and time-consuming task for most libraries. Therefore, it must be planned properly to complete the task alongside the library's day-to-day activities.

What is Retrospective Conversion?

Retrospective Conversion is the conversion of a library's existing bibliographic records from manual to machine-readable format according to specified policies and standards" - Jane Beaumont and Joseph P. Cox. Retrospective conversion is the process of converting a library's printed catalogue holdings into machine-readable form, in a pre-defined format using accepted standards for making it searchable through a machine. This process is a prerequisite for library computerization. Libraries cannot begin online transactions or other functions such as acquisition, cataloguing, circulation, PAC, and serial control, etc., unless and until retrospective conversion is fully done. Retrospective conversion is typically accomplished by one or a combination of two basic methods: matching and extracting machine-readable records from a resource file or keying records directly into a database. It is clear that retrospective conversion is concerned not with ongoing cataloguing, but rather with converting previously created bibliographic records from card or paper formats into machine-readable format. It is not cataloguing or even re-cataloguing. Although sometimes re-cataloguing may be necessary or useful to make changes to older headings, classification numbers, etc., the primary purpose is to create a machine-readable file of the existing manual catalogue, not to duplicate the work of previous cataloguers. It is the conversion of data from one format to another as quickly and as inexpensively as possible, maintaining the policies and standards set for the project. It is advisable to set the policies before starting the job, such as how much editing is to be done? Are all names to be checked in an authority file? etc. Standards for the task must also be decided upon, such as: will the Anglo-American Cataloguing Rules II (AACR-II) be followed faithfully? Is the machine-readable catalogue (MARC) format, CCF Format, or any other standard format to be used? It is a most troublesome task. Suffice it to say that for a project to be successfully accomplished, certain policies and standards must be decided upon before the project begins, and they must be adhered to during its execution.

Need for Retrospective Conversion

The conversion of a library's holdings database from non-machine-readable form to machine-readable form is a prerequisite to implementing an automated system. This database would become the foundation for other library activities such as online public access catalogue (OPAC), circulation, catalogue maintenance, resource sharing, etc. These records provide the means of generating statistics and other information needed to improve existing services and introduce new ones.

Retrospective Conversion Features/Benefits

To Create a Database for a Local Automation System: The library usually wishes to automate the circulation function in the library and/or to provide an online catalogue. The library may also wish to provide for its future automation needs and in fact proceeds to carry out a RECON project without having a specific automation system in place. The need to have a machine-readable file of its bibliographic records has, however, been identified.

To Improve Speed of Service to Users: It is to be expected that the results will improve the services that the library provides its users. Some of the benefits which library users may find are improved circulation services, improved access to the library's collection, improved interlibrary loan response time, etc.

To Improve Internal Library Procedures: Not only should a library expect a RECON project would benefit its users, it should also expect that the project would assist the library's own staff in accomplishing their tasks more efficiently. This may manifest itself in such benefits as the elimination of card filing, the availability of more current bibliographic data, the lessening of time-consuming changes to catalogue records, and an improvement in pre-order and pre-catalogue searching.

To Generate Special Products: If a library has its catalogue in machine-readable format, it is much easier to generate special products including a complete database produced in a different format (example, a catalogue on floppy diskette can be distributed at various points in the library or to different branches) or a subset of the database (example, a serials catalogue, or a catalogue of the library's non-book material, electronic media resources, etc.). Once the data have been entered into the library’s databases in a consistent format, with useful coding, it is possible to generate many special products. The library's catalogue is no longer site-bound and can be accessed more easily.

To Create an Integrated File and to Eliminate the Cost of Maintaining Parallel Systems: One of the reasons many libraries contemplate a RECON project is that present cataloguing is automated while older records remain in card format. This creates two files and hence two places to search for library materials. It is quite common for a library to have some of its records on floppy diskettes or in an online catalogue, and the remainder in a card catalogue. This is awkward for both users and library staff. With a complete RECON, the library acquires a single, integrated file. The cost of maintaining two catalogues has also become prohibitive for some libraries, and they simply close the card catalogues and maintain just one catalogue. Other libraries maintain both catalogues. Neither solution is very satisfactory. When the RECON project is completed, however, the need to maintain parallel catalogues is eliminated, and hence catalogue maintenance costs are reduced.
To Provide Flexibility in Changing Systems: It is evident that the rate of technological change is rapid. Some of the libraries that were the first to automate have now changed systems two or three times. A library must therefore try to create a database that will allow it to move with technological change. Having an integrated machine-readable file will allow libraries to move from older to newer computer systems more easily and less expensively.
To Maximize Return on Automation Expenditure: Libraries are spending vast amounts of money on computer hardware and software. If, for instance, only fifty percent of a library's catalogue is in the automated system, then the library is not maximizing the return on its investment. Besides not using its computer power to its full potential, the library is probably wasting time and money using and maintaining two systems, a manual one and an automated one.
To Protect Library Files: One of the attributes of machine-readable catalogues is their portability, as opposed to card catalogues, which are usually maintained at a single site. This portability provides added protection to the library's bibliographic files as copies of the files can be stored and maintained in off-site locations. In the event of a flood, fire, or other disasters in the library, the file copies remain safe. Preservation of the files should not pose a problem in the future, as they can be copied and transferred to newer and better means of storage as these become available.

To Create Union Lists: For libraries that are part of a larger library system or group of libraries, having all the holdings in machine-readable format can facilitate the production of union lists, which can be a cumbersome task if maintained manually.

To Share Bibliographic Data: The conversion of a library’s bibliographic records to machine-readable format means that the data can be shared with other libraries. This is especially true for those libraries that form a consortium or maintain files with bibliographic utilities. This sharing of data is mutually beneficial for the member libraries as the pool of available records increases.
To Share Bibliographic Resources: Not only is the bibliographic data made more widely available when in machine-readable format, the bibliographic items themselves are often more widely available to other libraries. Through the use of central databases and electronic mail, the interlibrary loan function in many libraries, especially smaller ones, has been greatly enhanced by the ability to identify a location for an item and subsequently to make an interlibrary loan request for it.
To Facilitate Collection Rationalization: The decision whether or not to purchase a specific item can be aided if the library in question has access to all bibliographic records of similar neighboring libraries. With automated union catalogues and bibliographic consortia, it is easy to explore the holdings of other libraries.
To Create National Databases and to Contribute to Universal Bibliographic Control: Having the library's records in machine-readable format can extend the benefits beyond the local and regional arenas to the national and international. When libraries contribute cataloguing to a larger pool, they are increasing the number of records available to all who use that pool; they are aiding the bibliographic control of items in their own area and contributing to universal bibliographic control.

Objectives

The objectives of retrospective conversion are:

To create a database for the automation system;
To maximize access to the collection;
To improve services;
Reduction in time for searching a document;
To improve library internal procedures: the integration of acquisition and cataloguing through automation and streamlining of other technical services; and
To maximize returns on automation expenditure.

Scope

To achieve the desired results, it is necessary to have a good understanding of the relationship of each aspect of the retrospective conversion process. Decisions made on one part of the project will have an impact on others. Thus, once it is decided to initiate the retrospective conversion project, it is essential to determine the scope of the project by:

Stages/phases and schedule of the process;
Deciding what areas of the collections will be converted;
Prioritizing the order of conversion;
Desired speed of conversion;
By whom and how the conversion is to be completed; and
Project costing and budget.

It is better to ensure that everyone involved in the retrospective conversion project understands the scope, goals, and objectives of the project. Staff participation is most important, non-cash input for the project's success.

Planning

For a successful retrospective conversion project, there is a need for a sound and detailed plan tempered with realistic expectations. The plan should:

Clearly identify the objectives;
Carefully document procedures (paying close attention to efficient workflow);
Specify standards to be used;
Identify reporting requirements;
Identify the necessary staff training; and
Plan for regular monitoring of quality and schedule.

If in-house conversion is to be done, each staff member on the project must clearly understand his/her tasks and responsibilities, the proper sequence of activities, and the standards to be enforced. If the conversion is undertaken through a vendor, care must be taken to ensure that the contract has no loopholes and that sufficient legal safeguards are included to protect the library in the event of serious difficulties with the vendor. In this regard, libraries can prevent many problems by requiring progress checkpoints and periodic tests of the quality of results.

Retrospective Conversion is never easy. It can be done successfully, with minimal problems, if expectations are clearly understood at the outset, and if planning and documentation precede the actual conversion. Once the project begins, success can be best assured through regular and frequent monitoring of progress and quality according to plan documents. The plan and schedule for the conversion project must take care of all the parameters of the conversion project. These parameters have multiple dependencies as follows:

Budget = f (number of records, fields, quality, and rate);
Quality = f (manpower quality, supervision, and planning);
Manpower = f (number of records and speed);
Manpower Quality = f (wages and training);
Supervision = f (quality and speed);
Speed = f (manpower, quality, number of fields);
Period = f (number of records, fields, speed).

If any of these variables is changed, it may affect several others. It implies that a single variable cannot be changed without affecting the related parameters. The above variables mean: Budget: Total budget of the conversion project. Quality: Standards and freedom from errors of records on OPAC. Manpower: Number of persons working on the conversion project.

The Planning Framework:

Important Questions The following considerations provide a logical progression and suggest a series of checkpoints to guide library administrators in developing a basic framework for retrospective conversion. First, librarians should examine four basic assumptions of library automation:

Provide patrons with a broad range of services in a timely, reliable, and cost-effective manner;
Allow the staff to complete necessary tasks with less effort, rather than more;
Requires a database of bibliographic records in order to function; and
The rate of expansion of library automation services is dependent on the allocation of resources and the growth of the bibliographic database. Given these assumptions, librarians should address these important questions:
What minimum set of records must be converted to establish a database that will support enhanced, labor-saving services;
What functions can be supported by library automation to achieve the desired level of services?; and
How should resources be allocated most economically to achieve a balance between retrospective conversion and the range of automated functions to be acquired? By working through these assumptions and questions, librarians can develop an approach to automation that links retrospective conversion and library automation so that these can be managed as a single integrated effort.

Pre-conversion Issues

Record Format: To ensure that the first conversion is the last conversion for the library or information center, it is important that the bibliographic information of the documents be converted into a standard format. Fields in the Database: It is time to decide which fields are to be included in the database and which are not. Current practices for recording bibliographic information may need modification when added to the machine-readable form. Decisions should be made after considering the users’ present requirements and future needs. Source for Bibliographical Information: Will the accession register, shelf list, or public catalogue be the source to record the bibliographic information for conversion? The choice should be based on the source that has accurate and complete information to meet the requirements and quality standards. Priorities of Conversion: Determine priorities for the areas of conversion so that more important areas of the holdings may be converted first into machine-readable format.

The Problems Behind Retrospective Conversion

Retrospective conversion is troublesome primarily because of the following facts:

With traditional techniques, retrospective conversion is very expensive;
Typically, it is a multi-year project; and
Since it is often considered such a lackluster or boring subject, the topic generates little managerial interest and involvement.

Despite the unfortunate reputation retrospective conversion has acquired, its importance cannot be overemphasized. The database resulting from a retrospective conversion project may long outlive the first, second, or third generation of automated systems installed in a library. Over the years, it can be very difficult to try to live with a poor database that is the result of budgetary or staff shortages. As Barbara Markus has commented, “Librarians never have enough money to do the job right the first time, but libraries have enough money to do the job over again a second time.”

Basic Retrospective Conversion Guidelines

The success of a conversion project is closely tied to a number of decisions. First, the library staff must decide whether to undertake a partial or full conversion and whether the conversion will be done in-house or through an outside agency. A library must make decisions of equally critical importance regarding the completeness and consistency of records. It is inevitable that converted records will differ qualitatively, to some extent, from new catalogues because converted records are derived from local data that are limited by previous cataloguing practices. Nevertheless, a library’s catalogue database is a long-term investment, and machine-readable records should reflect information that is as complete and accurate as possible.

Documentation

Well-written documentation unveils the project's strengths; conversely, poorly written documentation unravels its weaknesses, thus documentation is crucial for successful retrospective conversion and automation. If changes in procedures are necessary, add them to the specifications. To ensure an efficient workflow, a manual may be prepared for the members working on the project, which will lead to consistent practices in the conversion. 12. Retrospective Conversion Process At first sight, the process of developing a RECON project may seem rather onerous. For the project to be successful—that is, to be both cost-effective and efficient—forethought and planning are essential. It is necessary for a library to know clearly what it wants to achieve with the project, the options available for carrying it out, the cost implications of each option, and the most efficient way of proceeding with the selected option. The following eight areas are to be discussed before proceeding to the conversion of the data.

Determine Library Needs It is imperative for a library to determine its automation and machine-readable data needs. In this context, we will look at both present and future needs. With regard to present needs, the immediate goal of RECON must be established; the automated cataloguing methods already in place, if any, must be examined; and the overall automation situation of the library must be taken into account. With regard to future needs, the medium- and long-term plans for automation should be considered to determine the impact of such plans on the scope, standards, and quality of the Retrospective Conversion Project. Establish Parameters for the Recon Project Before meaningful parameters can be set for a project, it is necessary to know the content of the library's collection and how it is used. This includes knowing the overall extent of the collection, the subject and form divisions within the collection, and in general terms the use made of the various parts of the collection. This information is especially vital if only a portion of the collection can be converted at one time. Once this information is known, one, some, or all of the following parameters can be used to limit the scope of the project:

Circulation or use;
Data range;
Classification range;
Time frame;
Cost ceilings;
Reclassification; and
Re-cataloguing.

Establish Standards for the Recon Project Some of the most crucial decisions to be made in planning the RECON project are in the area of standards. To establish standards for the RECON project is to determine the level of acceptance by an individual library of generally accepted and widely used cataloging codes and machine-readable record format structures. While standards may be established by an agency outside a library, they are usually accepted or modified by the individual library and are determined by library needs. We must consider, therefore, how standards and library needs, both present and future, are dependent upon one another. While standards depend on needs, they also have a great impact on the cost of a RECON project. Therefore, the cost implications of local acceptance of certain standards must be considered. It is necessary in this discussion to examine the following standards:

Level of fullness of the cataloguing record;
The use of AACR2;
The use of the MARC format, CCF format, or any other format;
The use of subject access systems;
The use of classification systems; and
The use of authority control devices.

Identify Recon Option The options available to libraries for RECON are many and varied. Each one has advantages and disadvantages, and the library must consider its options carefully. We will identify options and consider the implications of RECON carried out:

In-house;
Capturing data by machine - optical character recognition, voice data input;
By an outside RECON vendor/commercial vendor;
A combination of in-house and vendor RECON; and
Retrospective conversion by OCLC - by using OCLC RETROCON services.

Analyze Options Once a library has identified some potential options for RECON, it is necessary to examine those options in more detail. This will allow the library to explore the impact of each option on library resources and to identify the qualities and characteristics of specific options. When analyzing options, a library will consider:

Library resources;
Library catalogues;
Library requirements;
Vendor characteristics; and
Vendor requirements.

In the analysis of options, it is obviously essential to look at costs. For each option chosen for analysis, it is necessary to determine the costs for the following:

Staff;
Equipment and supplies; and
Vendor or utility charges.

Determining the costs has proven difficult for many libraries involved in a RECON project as there are so many variables at work, and not all are fully known prior to the commencement of the project. Choose an Option Having gathered the necessary data as outlined above, a library is now in a position to decide which option to choose. The ideal situation would be to find the option that offers the lowest cost, the highest quality, and the one with the least negative impact on the library. However, what is likely to be chosen is a compromise between cost, quality, and impact.

Develop Plan of Action After choosing a method of RECON, it is time to get down to the details of how to carry out the project. Whether it is to be carried out in-house, by an outside vendor, or through a combination of the two, it is necessary to assign supervisory staff to the project, specifically someone who will be directly responsible for its execution. Following this, it is necessary to establish written policies for the project, detailed enough to allow RECON staff to proceed in a consistent manner. As well as written policies, detailed user manuals should be provided for staff, particularly if the staff are newly recruited for the project and are unfamiliar with library policies and procedures. Developing the plan of action also involves determining staffing requirements and initiating the recruitment process. Finally comes the task of allocating the necessary resources for the project, including staff, equipment and supplies, work location, and money. Implement the project, monitor progress, fine-tune procedures, and complete the project While the project is being carried out, it is likely that changes will have to be made to certain procedures. Flexibility should be allowed for in policies and procedures in order to advance the project to its ultimate conclusion.

Methods of Retrospective Conversion

There are a number of ways to convert the catalogue into machine-readable form. The best method for a library or information centre depends upon the available resources.

In-House Conversion using Existing Staff

In in-house conversion, the conversion is completed by the existing library staff, which leads to high quality and control, as the staff understands the users’ needs, quality requirements, and the objectives of the conversion well. But it has some disadvantages, such as:

It disturbs the routine work;
Increases workload;
More time is required for completion of the project;
Additional space and hardware requirements;
Added supervisory and quality-control efforts;
Increased personnel costs; and
May in fact make in-house conversions more costly in the long run.

Outsourced In-House Conversion

In outsourced in-house conversion, the conversion is completed by outside contracted persons within the library premises. It is easy to meet quality standards as conversion is done under the supervision of the library staff, and there is less negative impact on routine work and staff time. However,

The cost of conversion is very high;
Space must be found for the additional temporary personnel;
Able to work smoothly with existing staff;
Access to a database against which to match; and
Convert your records (as with any in-house conversion).

Outsourced Off-site Conversion and Machine-Match Conversion

In outsourced off-site conversion, the process is completed by an agency away from the library or information centre. The advantage of it is only that the process is completed within the timeframe with less impact on the library’s routine work. But it has the following disadvantages:

Conversion cost per item will be more;
The library has the least control during the process; and
The shelf list/catalogue may probably have to leave the library.

One possible solution to the above is to provide photocopies of the title page of each document, but it incurs additional cost and is labor-intensive. There is another method: each item in the collection—as described in the shelf list—is matched to a database owned by the vendor. Alternatively, a bibliographic database provided by the library, usually as an extraction from a non-MARC-based local system, may be processed by the vendor and then run against a MARC bibliographic file. While a keyed conversion will likely cost more per item and will require the shelf list to leave the library, a machine-matched conversion may be done in a very cost-effective manner. Keyed conversions will make use of photocopies of the title page and verso if there is no shelf list or if the library cannot bear to part with its shelf list. However, photocopying

Is extremely labor-intensive, and
Will incur additional costs for supplies.

Performance measures and quality control concerns must be contractually negotiated, since there will be little direct control by the library over these. However, contracting with an outside vendor will result in predefined costs and timeframes for completion. An off-site bureau conversion can therefore often be performed more efficiently with much less impact on a library's day-to-day operations.

In practice, a hybrid approach is often adopted in which all three methods are used. The bulk of the monograph collection may be sent to a vendor to be converted off-site. More difficult materials such as serials, non-print materials may be converted on-site using either existing staff or contract labor, or these materials may be sent to yet another retrospective conversion vendor that specializes in converting complex, unusual items.

Steps in Retrospective Conversion

Filling of Data Input Sheets/Worksheets: Current practices for recording bibliographic information may need modification when converted to machine-readable format. Thus, as per requirements, Data Input sheets/Worksheets may be printed so that all the required information may be noted, ensuring no field is forgotten and none repeated. Those who have good knowledge of cataloguing, the scope, goals, and objectives of the project must design the Data Input Sheets. The person employed to fill up the Data Input Sheets must have good knowledge of cataloguing and must be aware of what information has to be recorded and what not. These Data Input Sheets/Worksheets must be checked by Library Staff to ensure that correct and required information has been recorded to meet the quality standards.

Entering Data into Software: After filling the Data Input Sheets/Worksheets, the information is keyed into the Library Automation Package. This can be done by someone who has basic computer knowledge and is a good typist. Here, knowledge of cataloguing is not required, but it will be helpful if the person is made aware of cataloguing rules to reduce mistakes.

Editing of the Database: After keying the bibliographic information into the software, the database must be checked for errors and edited by a responsible person to achieve the quality standards. This process must be repeated in a loop until the required precision is achieved. Typographic and spelling mistakes make the catalogue problematic for searching.

Key to Success

The key to successful retrospective conversions is careful and meticulous planning. There are several actions a library can take that will contribute significantly to the success of the conversion project. These measures include:

Careful planning;
Close examination of each method of conversion, taking into account the library’s budget, time, and manpower constraints;
Weeding of titles and copies that have marginal value. Since it is expensive, it makes no economic sense to convert materials that are of little or no worth.

SUBHAM SAHU

MLIS Student, SAMBALPUR UNIVERSITY

References

^[1] ^[2] ^[3]

↑ http://ir.inflibnet.ac.in/handle/1944/187
↑ https://www.researchgate.net/publication/267479771_Retrospective_Conversion_Problems_and_Prospects
↑ ir.inflibnet.ac.in › bitstreamPDF Web results Retrospective Conversion - IR @ INFLIBNET