Biblio Tech
Review
Information Technology for Libraries

Main BTR Sections

[Biblio Tech Review]
[2001 Issues]
[Tech Briefings]
[Product Directory]
[Exhibitions]
[Archive Issues]
[Archive Reviews]
[About BTR]

Reviews

[PBS update]
[Data Magician]
[ILL Update 2000]
[PBM]
[Libero]
[Dynix WebPAC]
[TMQ]
[Bibliostat]
[Voyager]
[NOTEbookS]
[Z39.50 clients]

© Biblio Tech Ltd. All rights reserved.

Personal research assistants

Flexible organisation and manipulation of the records in the database is essential and PBMs have evolved the concepts of marking and grouping records for subsequent actions like printing or exporting.

Record importing

Each of the products reviewed allows the user to import records from files and electronic products of various kinds – so that the majority of references may not be manually entered.  Many CD-ROM products have output formats for PBMs to simplify the process.  The PBMs themselves allow the user to customise the import procedures to suit new sources.  A curious omission from most of the PBMs is the ability to import MARC records.  There are millions of MARC records out there but few library systems or other services considered exporting records as a user service.  Now that Z39.50 services are commonly available the MARC record has taken on new importance as a potential standard export format to such end-user tools as PBMs for books and reports.

Article scanning and reference formatting

A PBM should assist the user by allowing them to add references easily from the database into an article as it is created, by quick look up of the database and selection of the relevant article to insert. This feature is becoming known as Cite While You Write (CWYW) and, with the close integration of the PBM into the Word Processor via a tool bar, it is probably a more efficient and accurate way than the batch scanning of a document.

Batch scanning identifies references usually by brackets around key terms e.g. (Evans, 1998) and links them to the database automatically.  This works well if you know your citations by heart and can more quickly enter something like (Evans, 1998) than search the database.  Once the database gets quite big however, selecting from the database is likely to be the best method. The batch method will allow you to link old documents to the PBM provided a consistent and recognisable identification method has been used.

Formatting the references in different output styles is a key function of the PBM and where they compete quite strongly to gain advantages in the number of formats supported etc. It is also a feature that would be difficult for a general user to customise a general product like Access to do. It is important that formats can be created and modified by the user to cope with the inevitable journal formats that do not come pre-configured.

All the PBMs offer professional layout facilities for the references – they can be presented as a bibliography at the end of the document or in the text – replacing the reference typed in by the author – whatever is required by the publisher.  Separate options are usually available for creating output separate from an article containing citations – this might be an output file for loading into another database or a subject bibliography.  HTML output is now also common for direct loading of a bibliography to the Web.
Web Poster
There will be a future review of Reference Web Poster from RIS which enables a bibliography to be posted on an Intranet or WWW and output to be selected and output back to another PBM.

Co-operation

Much of research is co-operative with teams working on a single project. Control and use of associated bibliographic references also needs to be a team activity. Network capabilities for a PBM allow a single database to be shared by co-workers with file locking so that only one user can access the database at a time for updating, but many can access for read only purposes – although Reference Manager now has field level locking to enable two users to access the same record at the same time.  Reference Web Poster will allow some of the difficulties of networking to be bypassed.

Trends and the future

The Personal Bibliographical Manager has been around since the early days of DOS and the core functionality is well established.  Inevitably, the major influences on new features are the Web, Z39.50 and integration with desktop tools.  The modern PBM should now be able to link direct from their database to any Web or desktop resource via a hypertext link so that if the full text is stored somewhere accessible then the PBM becomes more than just the catalogue for the reference, it becomes the researcher’s entry point to the world of information resources – it is tempting to apply the concept of a Portal – common now on the Web to describe the first contact point for searching and Web based services – to a PBM on the desktop.  A “Personal Portal” perhaps..?

As well as being able to access the Web, Z39.50 capability is obviously important since it potentially means that a researcher can access a library catalogue anywhere in the world and download a record directly into the PBM. The use of MARC as an internationally agreed standard for format means that only one format needs to be programmed in order to accept records from any library – unlike the plethora of formats required for talking to CD-ROM and on-line services. There are two ways of approaching this – either by integrating the Z39.50 software directly into the product, as with Endnote, or by interfacing with a widely available tool such as BookWhere as is possible with Procite.

Learning from the PBM

The PBM has grown up almost entirely separate from the library automation industry – but it has something to teach the catalogue management modules within Library Management Systems. Anyone wanting ideas to develop and improve facilities for cataloguing modules should take a look at some of the sorting and display options, the global search and replace, the duplicate checking options and many other features in any PBM.  These are functions grown from a different set of user perspectives, but I bet there are many cataloguers who would love to have this level of flexibility in their library catalogue offerings.

Improvements to PBMs

Overall, the PBMs reviewed did a good job of their primary objective – to take the hard work out of citation control within an article – and they are very strong in printed output control.  There are improvements that could be introduced however – notably in the introduction of a proper thesaurus structure to their subject keywording with preferred and non-preferred term control.

Searching is difficult to make powerful and simple. The PBMs try each in their different ways but none succeeds completely in my view - see the individual reviews for the detail.

The PBM and the Library Management System

Personal bibliographic software is primarily used in an academic research environment where the big Library Management Systems (LMS) are likely to play the major role in organising the corporate knowledge resources (library).  The PBM needs to communicate well with the LMS to allow the researcher to maintain personal copies of catalogue records (albeit with the risk of being out of date as records are updated).  Communication can either be by a “push” from the library system to the PBM – exporting a record in the OPAC to a local file which can then be imported, or by “pull” where the PBM addresses the Library catalogue and extracts the record directly – e.g. via a Z39.50 desktop client.

Any well specified LMS should have at least one of these methods of communicating their records with the individual researcher – the best will have both.

Librarians using their own LMS cataloguing sub-systems have long bemoaned their limited output and formatting capabilities.  If the system can output records in comma delimited or tagged format then catalogue records can be imported into a PBM tool for customising the data ready for final printing. It can be a lot easier than trying to re-format using a Word Processor – especially give the flexibility of selecting and the pre-built styles. In fact with the size of disc on modern PCs downloading a large chunk or even the whole library catalogue is not unfeasible. Using Procite, 100,000 records would occupy about 350 MB of disc space – although I would hesitate to predict aspects like search response times and selection on non indexed fields.

Summary

The choice of PBM will hinge, as ever with software, on the user’s needs.  Each of the products has strengths and all do the core job well. Where they differ is in their style of interface – which can be a very personal thing, management of the database – searching etc. and some of the detailed functions and features which we have tried to highlight in the table of features. The different reviews expand on some of the key features – good and bad – but do not cover every aspect. These products are complex and some of the real devil for an individual researcher will be hidden in the detail of what can and cannot be done.

Features score

The traditional features score in software reviews can be misleading when each product has some features that the others do not.  The scores here do reflect an overall capability but do check the full features tables and the reviews for individual details.

Although some products are shown to NOT have a feature, it is often possible to achive this function by a "work around". For instance, Endnote does not have an explicit NOT function but the same effect can be achieved by subtracting one set of search results from another. Similarly although groups of records cannot be created easily - you can export records from Endnote and then import into another empty library to achieve a sort of grouping.