Review of Personal Bibliographic Systems
ILRT Research Report Number: 1032
Publication Date: 2003/06/09
Last Modified : 2003/06/09 17:32
Author: Paul
Shabajee
Abstract
This report is a review of personal bibliographic
management software and systems – it provides a fairly comprehensive
review of available systems, functionality and issues in personal
bibliographic data management. It was produced as part of background
research for the SWAD-E Semantic Blogging Demonstrator Research and was
funded by HP Labs as part of the SWARA project based in ILRT.
(This is an edited version of report originally published as Appendex
D. to SWAD-Europe deliverable 12.1.2: Semantic Blogging and Bibliographies
- Requirements Specification by Steve Cayzer and Paul Shabajee - http://www.w3.org/2001/sw/Europe/reports/requirements-demo-1/hp-requirements-specification.html)
Introduction
The personal and small group management and publication of
bibliographic information is a ubiquitous problem for academics,
researchers, students, writers and more broadly authors of many kinds.
Many systems have been developed by individuals, groups and organizations
to meet their own specific requirements, some these utilise software, that
supports the capture, management, sharing and publishing bibliographic
information.
This appendix aims to provide a brief overview of key issues in the
context of the creation of bibliographies, these are:
- the kind of motivations behind the personal and small group use of
bibliographic management systems - be they paper or electronic or a
hybrid
- the functionality required by users and offered by the various
computer based systems
- the software systems available
- relevant standards
As part of this work we reviewed 20 (see below) bibliographic software
products and services, reviewed on-line literature and conducted a small
interview study of users of such systems (see Appendix
C). The interview study consisted of 5 interviews, with individuals
from academic and commercial backgrounds, they were asked about their use
of bibliographic data, systems including software that they have used to
capture, manage and publish the data, along with any positive or negative
feedback about the systems. Finally they were asked about 'wish list'
items, that they would include in any ideal system.
Reasons for Creating and Using Personal and Small Group Bibliographies
bibliography: "a list of the books and articles that
have been used by someone when writing a particular book or article"
Cambridge Dictionaries Online (http://dictionary.cambridge.org/)
The definition above indicates that the term bibliography
relates to a particular list of books or articles. However in broader
definition might cover any type of 'work' or 'resource' and may simply be
a list of works held for any purpose.
There are a large number of specific reasons for capturing, managing,
sharing and publishing bibliographic data - for example: creation of
bibliographies; as part of a personal knowledge management toolset;
organizing a collection of books or papers; group annotation or indexing
of resources; locating resources previously used. Although no specific
research in this area has been identified as part of this study, it seems
likely (based on the functionality of software systems and small scale
interview survey), that the dominant uses are to produce bibliographies
related to specific topics. Specific examples of the need to create a
bibliography include: student essays, grant applications, curricula vitae,
book chapters, project reports, sharing references with colleagues, and
reading lists. In general these lists must be formatted in specific manner
depending on the standards of organization and/or publication.
Needs, Functionality and Affordances
In order for a user or a group of users to do any of the things listed
in the previous section, it is necessary for any system to facilitate a
set of core activities, these can be broken down in a number of ways, the
following is a fairly generic set:
- assisted capture of necessary bibliographic (and other e.g.
notes) data
- management and manipulation of the data once captured, for
example: editing of records; locating of previously entered data;
searching for subsets of data for the creation of a bibliography;
ability to categorise and annotate the records.
- publishing and sharing of the data in an appropriate
bibliographic format.
There will be specific needs related to each of these three categories
in any particular context. However if we focus on the examples related to
the creation of bibliographies, these include:
Assisted Capture
Users are likely to obtain data from a range of sources - extracting
the necessary bibliographic data from these can be problematic. For
example: it is time consuming; it is difficult to ensure accuracy and
consistency of terms (e.g. spelling and format of author names); it is
often difficult to identify all necessary data from or about a resource.
Thus functionality such as the ability to automate or semi-automate the
identification and capture of data are helpful.
Management and Manipulation
Once the data is captured it generally needs to be managed and
sometimes augmented. The requirements with respect to the creation of
bibliographies are less extensive than other potential applications (e.g.
knowledge management). However there are a great many functional
requirements even in the relatively simple context. Some specific examples
of tasks identified during our interview (see Appendix
C) survey include:
- users may wish to add extra keywords - requiring tools to assist
in consistent key-wording e.g. retrieve all records indexed with a
particular keyword and index/sub-divide them using finer grain terms
- editing records to ensure consistency (e.g. UK or US spellings)
- adding bibliographic data as it becomes available
- augmenting notes and annotations about a work following further
reading.
- the desire to understand relationships between works, e.g. by
drawing timelines or mindmaps from the data
Fundamental to these and others is the ability to retrieve records
efficiently and effectively. Coupled with this is the ability to annotate,
organise and in some cases visualise the data.
Publishing and Sharing
'Publishing' of the bibliographic data takes a number of forms. In
practice the most basic case is the creation of a simple formatted text
list (bibliography) of selected records drawn from the system. The
format/style (see below) and media (e.g. paper, word processor file, HTML)
of the list, will depend on the particular context. More complex examples
include, the automatic embedding of references and citations within a word
processor document (cite while you write, type system, see below).
The sharing of bibliographic data with individual colleagues or within
teams, is a common practice, e.g. passing on references to colleagues for
their use in their report or small teams of researchers keeping a common
word processor or bibliographic database file.
Sharing data is used for a number of purposes, those identified as part
of our study (see Appendix
C) included:
- prevent duplication of effort in finding references
- as part of authoring process e.g. academic papers
- flagging up newly found interesting documents to members of a team
- share thoughts about research papers or publications (e.g. via
notes fields)
- ensure common format for data captured
- helping members of a team keep track of new developments/research
in a research field
Standards
There are a variety of aspects of bibliographic data capture,
management and publishing systems that require standards to be set, if
systems and data are to be interoperable. These include, the pieces of
data to be captured, the naming of the database fields, the organization
of the data, the rules of when and how to use a particular indexing [or
keyword] terms, the syntax of the data within fields, and the storage
formats. Appendix
E reviews some of the more extensive library related standards.
While the library standards are largely 'overkill' with respect to
personal and small group management of data to produce a bibliography,
many of the basic needs behind the requirements remain the same. From Appendix
E:
- what information should be captured about the ‘publication’ i.e.
cataloguing data
- the structure of the record
- detailed rules or guidelines for how to deal with specific
cataloguing situations/issues e.g. what to do when there are two
different formats of an authors name used within one publication.
Broadly these also include the use of authority lists/files as
definitive authorities over for example the spelling of a place or
personal name.
- how the subject or content of the ‘publication’ should be
described i.e. how the publication should be indexed.
- the specific syntax (e.g. use of punctuation) for encoding of the
record
- The specific transfer protocols for transferring the data between
locations.
In these areas, the formal library standards are well defined. However
at a personal and small group level, they are largely poorly defined or
there are many competing 'standards'. Examples include citation styles and
interchange formats; these will be discussed in turn.
Citation Styles
There are a very large number of citation styles i.e. how citations are
written to acknowledge works used or referred to in a document, and how
the references should be formatted, e.g. one of the most commonly used is
the Harvard style in which the: author and year of publication is written
(i.e. cited) in the text (e.g. J. Smith (2002)) of a publication. In the
case of a book, the reference contains the name(s) of the author(s),
editor(s) or the institution responsible for writing the book, date of
publication (in brackets), title and subtitle (if any) should be
underlined or highlighted or in italics, but must be consistent throughout
the bibliography, series and individual volume number (if any), edition
(if not the first), place of publication (if known) and publisher. The
references are place in alphabetical order of the family name of the main
author. Other types of publication will have different data e.g. in the
case of a journal the journal volume and number.
The other basic citation method is the 'numeric system' (Vancouver)
style, in which references are cited by a number in the text which is then
used to label and order the references in the bibliography.
There are generic, high level, international standards, for example:
However with regard to specifics of the syntax and format of citation
and references traditionally different disciplines have tended to use
different styles e.g.
- Humanities: MLA (Modern Language Association) -
MLA Handbook for Writers of Research Papers, 5th ed. MLA, New York,
1999) & The Chicago Manual of Style,
- Scientific: APA (American Psychological
Association) - Publication Manual of the American Psychological
Association, 5th ed. APA, Washington: 2001 & CBE (Council of
Biology Editors) - Scientific Style and Format, 6th ed. Council of
Biology Editors, 1994)
- History: Turabian - A Manual for Writers of Term
Papers, Theses, and Dissertations 6th edition, Kate Turabian's,
University of Chicago Press, Chicago: 1996. and Chicago (as above)
In addition, many journals, publishers, governments, corporate bodies
and other organizations define their own 'house style'.
In many ways these style rules are analogous to the International
Standard Bibliographic Description, ISBD (http://www.ifla.org/VII/s13/pubs/isbd.htm)
and AACR2 ( Anglo American Cataloguing Rules http://www.nlc-bnc.ca/jsc/)
standards for larger systems, discussed in Appendix
E, in that they provide similar guidelines (e.g. data to be captured
and order, syntax etc.), but for a fewer pieces of data
(fields/elements).
Information Interchange Formats
There are no ubiquitously used standards to encode personal and small
group bibliographic data i.e. there is no equivalent of MARC (see Appendix
E) in the library world. In general the commercial software packages
specifically designed for individuals and small groups use proprietary
formats for internal storage and in many cases export options are very
restrictive e.g. EndNote (probably the dominant product for personal level
data in UK Higher Education) only has txt, rtf and HTML formats as options
under the 'export' menu.
In some cases it is possible to use the tools to output different
styles to create output 'styles' that match any other text based format.
e.g. in the case of EndNote (see table below) there is a fairly
comprehensive style template language. Base level formats such as comma
delimited text, can be imported as standard by nearly all of the packages
reviewed.
In contrast, the majority of such systems can import a great many
formats (including competitor's formats) and as in some cases advanced
users can create their own 'import filters' (i.e. parsing rules to extract
the necessary data from the source format). In general for example MARC
records can be imported and the data parsed to extract only the required
fields to match the internal data format, having done the necessary
mapping between data standards.
The Dublin Core [DC]
Metadata Initiative has a citation working group (http://dublincore.org/groups/citation/)
which is working on refinements and encoding of bibliographic data. Dublin
Core unqualified or qualified can provide a base level metadata standard.
There are also various 'cross walks' (mappings) between formats e.g. MARC
to Dublin Core (http://www.loc.gov/marc/marc2dc.html).
However as yet Dublin Core has not become widely used as the basis of a
data exchange system for commercial systems reviewed.
Other metadata and encoding related initiatives include the TEI (Text
Encoding Initiative - http://www.tei-c.org/), EAD (Encoded
Archival Description - http://www.loc.gov/ead/) and DocBook
(http://www.docbook.org/). These all
have elements that deal with bibliographic data, however again as these
are designed for different and more complex application areas they are not
used in the systems reviewed here.
BibTeX [BIBTEX]
is one of the more widely used (and older) formats that has been used and
well supported with publishing tools. BibTeX files are text files with
appropriate encoding - when printing a document using a LATEX
processor, the document contains markup that instructs the processor to
import and format the records appropriately. This facility means that the
BibTeX file can act as a data exchange format. And for example CiteSeer
[CITESEER],
a widely used digital document library, provides a BibTeX entry (as plain
text) which can be cut & pasted into a BibTeX file. BibTeX formatted
data can be imported into some of the commercial systems.
Software
The UK, academic market is dominated by ISI ResearchSoft (http://www.isiresearchsoft.com/)
who produce the EndNote, ProCite and Reference Manager products, all of
which are widely used in Higher Education institutions. However there are
a large number of pieces of software that are designed specifically to
facilitate the personal and small group capture, management, sharing and
publishing of bibliographic data and other generic applications that are
used to do the same (e.g. spreadsheets, word processors, and database
programs). The list below details the majority of the more developed
products and services, identified while researching this report.
The phrases in quotes marks are taken from the cited Web site and where
there are no comments from the authors, this means that the product had
the basic functionality (see end of this section), only 5 of the products
have been tried/tested in any depth, other data is taken from Web based
research.
| Archiva |
(http://www.legend2000.com/archiva/arc_index.asp?arcMenu=Archiva)
- "an advanced reference management system with integrated thesis
processor." |
|
AskSam
|
(http://www.asksam.com/)- a generic
data management system that is very flexible and facilitates the
capture, management and publishing of structured, semi-structured
and unstructured data. Thus just as with generic structured database
systems, it can be adapted to work with bibliographic data. |
|
Bibliographix |
(http://www.bibliographix.com/)
marketed as a publication planner as well as a bibliographic
management tool, it provides an 'ideas manager' in which ideas can
be indexed using a "short thesaurus" hence keeping and linking
related ideas. It also enables network access to group bibliographic
databases. |
| Biblioscape |
(http://www.biblioscape.com/)
There are different levels of product. The standard edition has all
the basic functionality and in addition allows relationships to be
defined between references, e.g. "Supportive", "Contradict" (it
calls this cross linking). There is a freeware version (http://www.biblioscape.com/biblioexpress.htm)
|
| Bookends |
(http://www.sonnysoftware.com/)
Is a Mac based product and is "a full-featured and cost-effective
bibliography/reference and information management system for
students and professionals" |
| Citation |
(http://www.citationonline.net/)
"Citation is a powerful and easy to use bibliographic database
system and notes organizer for research writing." |
| Database Software (Personal)
|
Generic database systems such
as Microsoft Access that individuals or groups use to create small
scale applications to manage their bibliographic data. Other
examples include: Cardbox: http://www.cardbox.co.uk/ and
Idealist (http://www.bekon.com/)
|
| EndNote |
(http://www.endnote.com/) EndNote
is from is from ISI ResearchSoft and seems to be the dominant
product for personal bibliographic data management in UK Higher
Education. It has all the basic features, including very extensive
and customisable import and export filters, and a 'cite while you
write' facility which integrates with MS Word and Word Perfect. It
provides means of importing images and other files. It can use
Reference Web Poster (http://www.endnote.com/rwprod.htm)
which enables the Web publishing (on their server) of EndNote
bibliographies. |
| Spreadsheet Software |
Spreadsheet software is can
be used to store bibliographic data as a flat field database - the
columns being the data fields. These generally providing sorting of
the table, fields can be added as required and more sophisticated
versions (e.g. MS Excel - http://www.microsoft.com/office/excel/)
provide filtering by columns. However there is no easy way to
produce actual bibliographies. |
| Library Master |
(http://www.balboa-software.com/)
"Library Master automatically formats the bibliography, footnotes
and citations for your paper, thesis or book in numerous
bibliographic styles. It makes it easy to organize research notes
and project records." |
| Ibidem/Nota Bene |
(http://www.notabene.com/brochure/ibidem.html)
"Store bibliographic information in the simple database format, and
Ibidem will generate your bibliographic references for you." |
| Papyrus |
(http://www.researchsoftwaredesign.com/)
Has DOS (version 7) that runs under Windows, and a Mac version
(version 8). In addition to basic functionality this provides
linking between records e.g. relationships such as "Reviews" or
"Refutes" and keywords e.g. "Synonyms",
"Supercategory/Subcategory,". Allows the embedding of images. It
claims to allow import of references form 'anywhere' including from
existing bibliographies in word processor format, using "artificial
intelligence techniques in reading your source file, alerting you to
potential problems" |
| PowerRef for Windows |
(http://www.cheminnovation.com/powerref.html)
provides the majority of basic features along with high levels of
flexibility in document type, adding user defined fields also allows
attaching of graphics to records. |
| ProCite |
(http://www.procite.com/) ProCite
is also from ISI ResearchSoft and provides comprehensive basic
functionality plus a network version with access to a single file,
multiple read but only one person can write at one time. It also
captures the URL a Web pages title information and stores in the
reference collection, and then text from the page can be pasted
directly into the ProCite record - although one interviewee noted
problems using this facility. |
| Pybliographer |
(http://canvas.gnome.org:65348/pybliographer/)
is a Linux based product, it has a basic level of functionality.
Licensing is based on a GNU GENERAL PUBLIC LICENSE i.e. "Everyone is
permitted to copy and distribute verbatim copies of this license
document, but changing it is not allowed." (http://www.gnu.org/copyleft/gpl.html) |
| Reference Manager |
(http://www.refman.com/) Also from
ISI ResearchSoft, again with comprehensive basic functionality along
with full multi-user networking i.e. multiple uses can both read and
write to the shared database. |
| RefWorks |
(http://www.refworks.com/ This
was the more extensive of the two Internet-based services that we
found. It is designed specifically for group as well as individual
use. |
| Scholar's Aid |
(http://www.scholarsaid.com/)
"Scholar's Aid 2000 is a program package that includes a
bibliographical data manager called Library and a notes/information
manager called Notes", it seems to have a good range of basic
functionality. It also claims to export to XML format. |
| Scribe SA |
(http://www.scribesa.com/) Is a
online service designed to capture data via a forms interface - it
has different version for different output formats i.e. APA, MLA or
ISO 690. |
| WebCrimson |
(http://www.webcrimson.com/)
WebCrimson is essentially an online web publishing service. It uses
a server based content management system, which has a large number
of pre-defined data fields, to generate pages from templates, the
user can choose from a variety of predefined template and edit them
or create their own. They then add their data to the database via a
forms interface and the system generates the pages. One of the
pre-defined (customisable) templates is a bibliography template,
which (with some tweaking) produces Web-based
bibliographies. |
There are a number of other products (e.g. InMagic: http://www.inmagic.com/) and systems
(e.g. ADLIB: http://www.uk.adlibsoft.com/) that
are designed for a larger scale of usage.
As noted above there are very many products - commercial, open-source
and developed by individuals - of which this list is a relatively small
sample. However, we believe that the major products have been covered. We
are not aware of any publicly available data regarding market share of
products and/or actual usage of the systems. Anecdotal evidence in the UK
points towards Endnote being a dominant product in Higher Education.
In general the systems have similar basic functionality e.g.
template based data entry, simple search facility, import from Z39.50
servers, default and customisable output bibliography templates (see
below), in many cases integration with word processors, etc. Some also
provide support capture of electronic document and/or ideas
management/capture along with other functionality (see below) to provide
an integrated workspace for researchers, author, students and other
potential users.
Key Features
From the above review and our interview study we have complied a list
of 'features' that the various systems provide (with respect to
bibliographic data management, we have not included below other features
such as ideas management). These have been grouped into three sections;
data collection/entry, data management/manipulation & sharing,
publication/export. However in many cases features cross these areas or
choices in one impact on another; we have tried to make these clear in the
text.
Before going into detail about particular features it is worth
discussing issues related to platform/operating systems. In the majority
of cases the systems above are pieces of software installed on individual
machines (as opposed to Web-based products, such as RefWorks) The majority
of the packages above are Windows products, those which are Mac based or
have Mac versions include Bookends, EndNote, PAPYRUS Version 8.0 and
ProCite. The only Linux product that we found was Pybliographer.
The feature list below aims to provide a list of the many of the key
features (i.e. it is not comprehensive) that are available in the systems
reviewed. Some are very common; where features may only be available in a
small number of products, this has been noted.
Data Collection/Entry
Under this heading we include all aspects of the capture of
bibliographic data from any source.
- Pre-formatted publication types: different types of
publication require that different data is captured e.g. a journal is
likely to require 'journal title' and 'volume number', which are not
relevant to other types of publication. Most systems provided a number
of standard types along with the ability to add others and in some cases
edit the standard version e.g. to add a URL to a book reference.
- Specifically designed forms based input: to facilitate the
data entry e.g. most systems order the fields with author(s), title,
year and other key elements at the beginning of the form - the majority
do not allow customisation of this ordering e.g. one participant in our
survey felt that it was annoying to have to go to the end of a form to
enter the URL.
- Predictive and auto-completion: Some systems
monitor text as it is entered into a field, and if there is a matching
value in the database already it fills in the field, the user can accept
the value or keep typing. This can increase the speed of entry
considerably, however where many values are similar e.g. journal names,
this is less valuable, some may find it annoying.
- Indexing/Key-wording records: Many systems provide
means of applying keywords to documents to allow the categorising
/grouping of records. In general these were un-structured, one product
enabled users to link keywords via relationships e.g. e.g. "Synonyms,",
"Supercategory", "Subcategory,"
- Authority Files/Drop Down options: One participant
in our interview study had created their own database using Microsoft
Access, they had created separate tables for authors, journal names and
publishers (but not keyword) and added dropdown option lists on their
main data entry form to make data entry faster and more consistent.
While this is done to some degree by the predictive completion of text
(see above) the authority file is a more robust system. None of the
systems we actually tested had this 'drop down' type facility.
- Spell checking: related to authority files is the
facility to check spellings of all the text in a record, this was
provided in a many of the products.
- Import filters: these allow the user to import
bibliographic data held in file formats different to that of native
system, e.g. BibTeX, or data from competing products or CD Rom based
databases. Some products provide a simple parsing language to allow user
defined filters.
- Ability to search and retrieve records from remote Internet
accessible bibliographic database services: This is via the
Z39.50 protocol. Many systems provide a set of standard 'connection
files' for commonly used services, these are pre-configured to access
these. In general these can be edited and new ones created as necessary
- however this requires some work and generally some technical
expertise.
- Automatic field completion for Web pages: One
product had the facility to gather data directly from a Web browser and
so auto-complete some of the relevant fields and support capture of text
from the page into relevant fields.
- Document capture: Electronic documents themselves
(as opposed to simply the bibliographic details) can be captured e.g.
PDF and HTML files. These can then be used to do more advanced searching
e.g. full text searching.
- Attaching image objects to a record: Some products
allowed images to be attached to the record and stored in the internal
database.
Data Management/Manipulation & Sharing
This includes any processing, management, manipulation or sharing
taking place between capture and publishing, including sharing of
data.
- Searching/Filtering of records: Probably the most
basic requirement, most systems provide tools to search the data using
any field and in some cases multiple fields. Results can generally be
exported to an external bibliography.
- Customized/sorted views: The data can be sorted in
to different orders e.g. by author or title and in some cases a table
(spreadsheet) view. In some cases it is possible to define which fields
to view.
- Defining relationships between records: A small
number of the products offered the facility to define relationships
between records of the form "Supportive", "Contradict". These can then
be traversed to explore the records and select them for exporting into a
bibliography. Other systems allow references to be grouped, which is
similar in functional terms to keywording and filtering on specific
keywords.
- Hyperlinking to URLs: Many products provided URL
fields which provide a means of direct linking to the source document.
- Network and Multiple User Access: A number of the
products are designed to be used by groups. These generally use a server
based database or a single database on a network share. These take care
of locking the records to prevent simultaneous editing. Two of those
reviewed were Web based services in which the data was held on an
external server with logon from any Web browser. Of course many other
products can be used for non-simultaneous access to a database file on a
network share.
- Duplicate detection: One product provides duplicate
detection, hence assisting in maintaining the database, others provide
tools that could facilitate this.
- Global Search and Replace: Some products provide a
means of changing a piece of text across all records e.g. a name spelt
consistently incorrectly.
- Saving and Backing up: Many systems had database
back ends and so individual edits are automatically saved, however in
many cases actual backing up of the whole database, requires technical
knowledge e.g. where the files and which require backing up.
- Merging of database: In many cases a product
allowed the use of multiple databases, some provide tools to assist in
merging these, however there may be some problems with dealing with
duplicate and near duplicate records.
Publication/Export
- Creation of Bibliographies in commonly used styles:
This is probably the most basic publishing feature. The
majority of products allowed customisation of the templates or filters
using custom languages or integrated tools.
- Export filters: More generic than bibliography
publishing, is the ability to export the database or a selected sub-set
of the database to an external format. Most products offered very
limited export facilities, compared to their import features. This may
be to try and 'tie' customers into their product. The most basic format
seems to be standard comma delimited, with a custom ordering of fields.
- Integration with standard word processors: Many
products provide integration with word processors (e.g. MS Word and
WordPerfect). In most cases this takes the form of the ability to place
a mark (automatically or not) denoting a citation in the document. The
program then either automatically adds a reference to the references
section at the end of the document, or the document is processed later
to do so.
- Web posting/publishing of bibliography and/or
database: A very common facility was to output bibliographies
or the data in HTML or to an online and searchable database.
- Word processor templates for manuscripts of common
journals: Using Microsoft Word Macros one product leads the
user through the creation of the document with relevant formatting, text
styles and citation styles.
Usability
We are unaware of any usability studies with respect to the
bibliographic software reviewed. However from our limited review of
systems and comments gained during our interview study, it seems likely
that there are a number of usability barriers to the easy and continued
use of these systems in general and specific problems with individual
products.
Central to the problems we have identified seem to be ease of data
entry, all participants in our (albeit small) interview study indicated
that the time consuming and often 'fiddly' nature of data entry was
problematic. Other key issues relate to customization (e.g. difficulty in
creating customized styles), integration with other packages (e.g.
technical bugs) and the time consuming nature of simple day to day
maintenance. The bottom line for all bar one of our interviewees was that
they tended to use the software for personal use only when they need to
write papers or other documents with bibliographies, rather than as a
general repository for bibliographic data.
It would be interesting and useful to conduct, or find existing data
about, usability studies in this area.
Wish Lists
This 'wish list' of features are compiled from the interviewees in our
small scale study, and so can only be indicative of such items that might
come from a larger survey. It should be noted that in some cases the
features are available in some products, however they were not in the
system(s) used by the interviewees. These are ordered under rough
headings:
Capture:
- Auto extraction of data from text citations and references in
papers, for example the Windows clip board, including BibTeX format.
- Auto capture of bibliographic data from a paper e.g. where the
papers were from a common journal and thus had a common format - import
filter for whole papers.
- Help in 'getting into' a subject area e.g. the common problem of
knowing what to ask and what vocabulary to use in a new field. Possibly
have a system to cross search other peoples' data and say that 'people
who have indexed this paper have used the following terms' or 'expert y
has used the following terms to index this paper'.
- Links to external databases to confirm spellings of names, journal
titles etc.
- Import data from 'table of contents' services data inc. abstracts.
- Copy some/all details from one record to another e.g. when adding
lots of papers by the same author of same journal
- Capture references at the end of an electronic publication
- Auto or semi-auto detection of document type (e.g. book, journal,
etc)
- Notification of additions to selected internal and external
databases
Management & Augmentation:
- More effective classification tools inc. more structured
vocabularies
- Access and update [the] data online, from any machine via a browser
- Hierarchal keyword structures, "so that they are easier to navigate
and conceptualise"
- Richer annotation possibilities other than 'notes' fields and better
searching of them.
- Semi-automated indexing or keywording from the text of an on-line
document
- Visualization of aspects of data e.g. timelines for papers under a
given keyword, citation relationships, or identification of 'camps' of
researchers with particular points of view...
- More flexible merging of databases.
- Semi-automatic up dating of documents that have used a record [e.g.
in a bibliography] when the record changes - i.e. the system should
remember when and which files have used the individual records and
notify the user when changes are made
- More effective backup
Sharing:
- More extensive export facilities
- See or be notified, when a record was added or edited and by who
- Ability to select records and e-mail selection from within system
- Ability to search other people's bibliographic databases e.g. known
expert/specialist or researcher in field, in particular found via
publications e.g. person x published paper on y, therefore they might
have a bibliographic database.
Publication:
- Output to any bibliographic software
- Ability to integrate (export to and import from) and output to other
applications e.g. graphics, mind mapping, database or visualization
program, e.g. export data to a mind mapping program so that main
branches were authors, next level publication year and then title
- Easy selection of records for inclusion in a bibliography e.g. tick
boxes next to record on screen, multiple select (using ctrl, left mouse
button), shopping cart type approach
- XML output and input
While these are necessarily very extensive or generalisable to the
wider community of users and potential users of such systems, they do give
some insight into the perceived needs of users and the significant
potential for improvements in design, and functionality of the existing
software and services.
Sources of Reference
These sources were the main Web sites from which generic background
information was gained as part of our on-line literature review. In the
majority of cases information about products and services came from the
publishing organisations' Web sites. Other citations are referenced
directly in the text.
Evans, Peter. (2002) A review of 3 major Personal Bibliographic
Management tools. Available: http://www.biblio-tech.com/html/pbms.html.
Information Systems and Technology University of Waterloo. (2000) Which
Personal Bibliographic Management Package Should I Choose? Available: http://ist.uwaterloo.ca/ew/biblio/which.html.
Kent, T. (2002) Bibliographic Software. The UK Online User Group.
Available: http://www.ukolug.org.uk/links/biblio.htm.
Maggie. Shapland. (1999) Evaluation of Reference Management Software on
NT. 2001. University of Bristol. Available: http://www.cse.bris.ac.uk/~ccmjs/rmeval99.htm.
Memorial University of Newfoundland Libraries, Bibliographic Control
Services (2002) Technical standards for electronic bibliographic
data/metadata. Available: http://www.mun.ca/library/cat/standards.htm.
Morton, D. (2001) Personal Bibliography Software. Available: http://library.uwaterloo.ca/~dhmorton/dnh4.html.
Online Computer Library Center. (2003) OCLC Online Computer Library
Center Homepage. Available: http://www.oclc.org/home/.
