 |
Abstract:
In recent years Monash has been developing and extending its virtual library services with a number of in-house projects. When a need arose for a union list of Monash Library electronic resources a group of library staff developed the Electronic Resources Directory. This paper outlines the design and development of the Directory, a web-based system which draws on and extends the library's catalogue records for these materials.
Introduction
"Monash will develop a virtual library service to deliver scholarly information in digital, print and other forms to staff and students of the University, wherever they are located. Recognising the University's strategic direction of increasing involvement in distance education and multi-mode delivery, the Library will develop new ways of supporting students so that the quality of the Library's services will be a major attraction for intending students and researchers."[1]
In recent years Monash has been developing and extending its virtual library services with a number of in-house projects. These include the exam papers database[2], Web-Sesame[3], Electronic Reserve[4], video on demand and audio on demand[5], electronic Document Delivery requests[6] and unmediated Inter-Library Loan[7]. Coupled with the acquisition of database services such as Ovid[8] and ERL they have changed the nature of many library services from largely in-library-based to accessible-from-anywhere. During this period the library was developing policies regarding the acquisition and cataloguing of electronic resources, particularly serials.
By mid 1996 the library catalogue contained MARC records for over a hundred and fifty electronic resources with the prospect of many more. These records were almost impossible to find by someone looking specifically for an electronic solution because there were so few of them in a catalogue of over a million records, and there was no single search to limit to electronic resources. Branch libraries had home-pages and were manually producing lists of electronic resources, including those listed in the catalogue. The library had an outdated union list of CD-ROM databases which it was loathe to update as the format was unsuitable for placing on the web.
Although a relatively small part of the library's total collection offerings at present, electronic products represent a significant expenditure. Maximising access and utilisation makes good sense. An increasing number of Monash users never set foot on a Monash campus and for these users electronic resources may represent an excellent starting point for their library research.
In July 1996 Deidre Lowe was commissioned to investigate the management of electronic resources at Monash library and to recommend solutions for consistent processing and presentation, including the possibility of a union list of electronic resources in addition to the library catalogue. In September Deidre presented her report in the form of an issues paper [9] including a recommendation to develop the "Virtual Shelf", a database of electronic resources at Monash library. This paper describes the development of the Electronic Resources Directory (ERD)[10] which is an outcome of the Virtual Shelf issues paper9. It also examines the process involved in the Directory's development because it is an example of the excellent results which can be achieved through co-operation between various groups within the library.
Development Phase
A small group of interested staff began looking at the Virtual Shelf project in late 1996. They included representatives from Reader Services, Technical Services and Systems Support. The initial brief was to develop a prototype Virtual Shelf database within a minimal timeframe and within the following constraints:
Once the prototype was in place the group was able to comprehend the practical problems involved both from the users and providers perspective. It had been extremely difficult to do this previously. Technical Services staff involved in the project were motivated to change call numbers quickly and within a few weeks there were about a hundred records ready for testing. It was immediately obvious that creation of a database would add an extra processing step. Catalogue records would be processed into a database, then database records processed into web pages. What was really needed was a means for staff to enter additional information about various items and a way to link the non-catalogue "extras" to the MARC records as they were processed. The bibliographic record number was chosen as the primary key to be used throughout the Directory. This is a unique ten-digit number automatically assigned to each new bibliographic record.
The second phase of the project involved a larger prototype and development of automated processing of MARC records into Virtual Shelf records. This was necessary for two reasons:
The input mechanisms for the non-MARC data were then developed. The alternative was time-consuming manual editing of data into the records so it was a logical step to make web-forms and an develop an underlying file structure and cgi script. A simple web form, relating to a specific part of the Directory record, was developed for each group charged with entering Directory data. The forms were placed on the library's intranet and required staff to enter their library system password before they could add or edit data.
Figure 1 shows the Technical Services data-input form. Technical Services staff must add a Resource Type indicator and one or more broad subject Descriptors (from a list). Branch staff can add local notes to each record if they wish. This is where the value-added information such as links to helpfiles, additional information about the resource, information on acquiring passwords etc is added. If necessary a record can have a local note for each branch, and the note is unique to the branch and not able to be edited by another branch. Subject Librarians may add a brief note to a record if they wish. This note displays as a comment in the title-level web pages. This facility was added because a similar facility exists in our CDROM menus. It allows an explanation such as "for Monash users only" or a very brief description of what a title really contains if the title itself is uninformative.
|
|
It was decided to use our local 259 MARC tag[14] to contain link information for electronic resources. The 856 US MARC tag was considered unsuitable because there could be many 856 tags in an ABN-sourced record and there was no way for Monash to order the 856 display or indicate which link was the most suitable. As Web-Sesame could hotlink a URL from any field, the exact placement of URLs was not a major issue. Recent changes[15] have resulted in a review of the use of 259 or 856 MARC tags. It is likely the 856 will be used in future as this would allow the display of multiple links for some items. Multiple links are necessary because there is more than one access method, indeed more than one source available to Monash users, for many online resources.
Subject-based lists were recommend[16]. Library staff were keen to see this part of the project materialise as they felt it was one of the more valuable products. The problem lay in finding a suitable subject-form or thesaurus to follow. Initially the group developed a few very simple descriptors to demonstrate proof of concept. Once the prototype was in place it was publicised among library staff. The group received many constructive criticisms of the Directory, and the structure and layout of the directory pages, and made many changes as a result.
By early 1997 a regular updating mechanism was in place and the descriptor problem was being addressed. The group felt it was not the appropriate body to make this decision and asked for input from Subject Librarians and in particular the library's User Education and Reference sub-committee (UER). UER decided on a faculty and department-based approach with descriptors loosely following the university's deparmental breakdown. This resulted in about 125 descriptors in a hierarchy[17]. A library-wide top-level menu-structure was agreed to. This menu-structure was used as the top level of the Directory and on the library's CD-ROM menus (eventually these two menu-systems will be merged).
The top level Directory page was immediately changed to match this structure, and the new-look Electronic Resources Directory was announced within the library. There was renewed interest and more suggestions for improvements followed. One staff member who was not part of the original group spent a great deal of time gathering feedback from other staff and even more time producing some graphics to make the pages look more professional.
The Directory then consisted of a series of html files describing individual resources and lists relating to each top-level subject-descriptor. These files were updated several times a week. There were now about two hundred records and there was concern that the lists would become too long. It was always intended to have the ability to produce lists at individual descriptor levels, or at higher levels, and to have hierarchical links between lists. This was one of the ways to add value to the information, by making it easier to browse suitable sources. Although the top-level menu structure was fixed, all levels beneath that needed to be fluid. The ability to add or subtract lists, and to add or remove particular descriptors from a list contents at will was required. This was achieved with the development of a configuration file, maintained by Technical Services staff, of the lists to be produced and the descriptors and subsidiary lists (if any) to be linked. In early October 1997 UER asked that a list be produced for each descriptor and these be linked from the appropriate top-level lists. The configuration file was edited, and the structure of the Directory changed as of the next update run.
Design Issues
Use of the Directory is intended to be intuitive. This is not easy to achieve.[18] Monash Library users can be anywhere in the world, English may not be their first language and some may have little or no contact with library staff. They may be using any of dozens of combinations of hardware and software. The group has tried to keep the screens simple, uncluttered and with clear text-based navigation buttons and icons. The web pages have not been developed with any one browser in mind.
Early versions of the Directory (the term Virtual Shelf was discarded because many Subject Librarians thought it would be relatively meaningless to library users) contained alphabetical lists of resources. Before they could connect to a source, users needed to view the record for a title then select a hotlink, if one existed. This closely resembled the structure of an online catalogue search.
The individual records also tended to resemble a catalogue record. Initially the records included call numbers for paper versions of items. These were removed because they took up valuable top-of-screen real-estate and were not really relevant to an online resources directory.
|
Figure 2 shows the Directory record for the electronic journal Aquatic Sciences. There are only so many ways to display bibliographic information, and the Directory record still resembles a library catalogue record in many ways. The title, connect information, notes and subjects come from the MARC record. The rest is additional data input by library staff. There are hotlinks to descriptor lists to facilitate subject browsing. There is a two-way link to and from the Web-Sesame record for the item, and links to a Directory search interface and Web-Sesame searches. For most records the most important information is displayed on the first screen.
The look and feel of individual Directory records has changed over time as library staff and users provided constructive criticism. The current iteration may change at a later stage, given sound design reasons.
In early 1997 the pages were changed to include a hotlink from the title lists as well as from the individual Directory records. This type of list went through several incarnations. The initial designs weren't intuitive to many users. The final design with simple text-buttons was well-received.
|
Figure 3 shows the Directory subject listing for Life Sciences. The "i" buttons resemble standard blue information indicators and link to individual Directory records. The "GO" buttons are green to resemble a traffic light and link directly to the resource. Once these symbols were chosen, the rest of the colours and images were coordinated around the blue and green theme. The intention was for library users to treat the Directory as a quick reference tool. In many cases the "GO" link should be sufficient. This has been borne out by the web server statistics. The pages which are most used are the subject lists and the search facility. The Directory records which are most looked at are the ones which lack "GO" buttons.
Links to subsidiary lists are placed at the top with a hotlink and a "down" arrow. Lower-level lists are similar, but with links to higher-level lists and an "up" arrow. Various formats were tried but this seems the simplest and most intuitive. Within subject-based lists, resources are divided by type. The current types used are: databases, electronic books, electronic journals, web-sites.
Limited user trials[19] were conducted. Users indicated a preference for the current version, finding it easy to understand and operate. Surprisingly some of the users who trialed the Directory were using text-only browsers, or browsers with images turned off. In general, users with non-image browsers were those connecting from overseas. Some thought had been given to non-image use and this proved to be sufficient. Two overseas users were extremely complimentary of the text-only provisions.
|
Figure 4 shows a complete Directory screen. The left-hand side of the screen contains a frame with the menu-items (the top level subject descriptors). This remains on screen during all Directory activities. It disappears if a user connects to resource. The frame is designed to fit well into an 800x600 screen (the Monash standard for student-laboratories). The right-hand side of the screen contains the lists, directory entries or search screen depending on the option selected. A frames solution was chosen because it allows the menu options to remain static. All non-Directory selections break out of the frame, preventing annoying "frames within frames within frames" displays.
Enhancements & Future Directions
The Directory is only as good as the data it contains. The Biomedical Library was the first to fully embrace the Electronic Resources Directory concept and add branch notes to all items of interest. This allowed the production of a branch-based list - one of the initial aims, but almost the last to be adopted. Branches can elect to add full notes or hypertext comments expressing interest in a title (solely for the production of lists).
A search-facility was an obvious enhancement. This was first added in early 1997. Initially it used a locally-written web-search feature which was rather limited but did not require any indexing. In July 1997 htDig[20] was installed. This provided rapid indexing and presented search-results in a format similar to many large Internet indexing services.
Another obvious enhancement is web-access to CD-ROM databases. This has been partially implemented. There is experimental access for library staff to some networked CD-ROMS. There has been considerable progress in this area and it should be available Monash-wide for the start of 1998. At this point Monash Library will have the option to move away from groups of specialised workstations to more generic research-workstations.
Some users, particularly library staff, felt that a new-items list, similar to the weekly lists produced for branch libraries, would be a useful addition. This feature was added in April 1997. The four most recent "new items added" lists are available at any time.
Cataloguing is a time-consuming and expensive process. The Directory affords an opportunity to include non-MARC files, such as lists of titles included in various CAUL trials. At the moment these titles are only available via the Directory search facility. If more detailed information becomes available, it can be incorporated as full Directory records. This would involve some programming work to format the data as html. Given that the total programming time (to date) spent on the Directory is less than 40 person-hours, incorporating a new dataset should not be prohibitive.
The full alphabetical and resource-type lists are very long (almost 700 titles to date). When there is time, they will be split into smaller lists. An "all subjects" list is also required so that users can browse the hierarchy, selecting a topic of interest. The images currently in use on the Directory were developed in-house. The group has a small budget for graphic design. It is likely that a graphic-design student will be engaged during the 1997/98 long vacation and a new look may result.
Reflections
The Electronic Resources Directory project has achieved its targets. It provides a flexible one-stop approach for browsing and searching electronic resources. The library catalogue is still the primary source of bibliographic information, however the changes made to accommodate the Directory have served to enhance the online catalogue as well, especially Web-Sesame searches.
Three of the original project group members are still actively involved with Directory promotion, development and maintenance, one other is still keenly interested in all related cataloguing matters. A fourth staff member is now actively involved with the group as well, promoting the Directory among Reader Services staff, providing graphics and many many positive suggestions. In several branches one person has become responsible for developing and maintaining branch-based notes. People and branches involved with the Directory tend to recommend additional resources for cataloguing, further enhancing the directory.
There is no doubt that Monash Library users find web-based services useful. Web-Sesame currently accounts for almost 29%[21] of total library-system use (that is all library transactions both OPAC and staff). The small number of Web-Sesame terminals in the libraries are popular. The bulk of non-library access is now via the web. Web-interfaces to services such as OVID have also proven popular. The Directory provides another service in a format users are coming to expect. It, or a successor, will eventually serve as the basis of a true one-stop front-end to all online library services.
[1]Monash University. Leading the Way. 1997. 28.
[2] http://www.lib.monash.edu.au/exams/
[3] Web-sesame is a locally-written front-end to the PALS online catalogue: http://www.lib.monash.edu.au/wwwlib/
[4] http://er01berwick.lib.monash.edu.au/library/
[5] http://audio.monash.edu.au/
[6] http://www.lib.monash.edu.au/docdel/
[7] MEADS is an unmediated-ILL project for STM library users. It is one of several unmediated ILL projects at Monash library. http://www.lib.monash.edu.au/meads/
[8] http://www.lib.monash.edu.au/biomed/ovid2.htm
[9] Lowe. 1996. 13-20.
[10] http://www.lib.monash.edu.au/er/
[11] Lowe. 1996. 13.
[12] Creating the initial 10 record prototype took about 16 person-hours . Creating the initial automated processing (which was in fact a complete processing system) took about 18 person-hours.
[13] New titles lists are created from a weekly report run: http://www.lib.monash.edu.au/newbooks/
[14] The 259 tag is used to display local serial holdings information
[15] Monash library was taking regular bib change files from ABN and local system constraints prevented any local changes to records as changes disappeared each time an update was applied. Changes in processing and system improvements in mid-1997 allowed the library to stop taking bib change tapes and consider other changes in MARC record maintenance procedures.
[16] Lowe. 1996. 18.
[17] The descriptor hierarchy can be viewed at http://www.monash.edu.au/journals/er/structure.html
[18] Hildreth. 1995.
[19] There was no time or budget for full trials. At various stages of development the author asked small groups of users to use the ERD and provide feedback. To avoid prior knowledge, a new group was asked each time.
[20] Htdig was developed as an intranet-index solution. It is an efficient index and search engine for a single site. See http://htdig.sdsu.edu/
[21] The usage figure is based on September/October system statistics.
Hildreth, Charles R. "The GUI OPAC: Approach with caution." The public-access computer systems review [online]. 1995, vol. 6, no. 5 [cited 1997-10-06]. Available from World Wide Web: <http://info.lib.uh.edu/pr/v6/n5/cont6n5.html>. ISSN 1048-6542.
Lowe, Deidre. "The Virtual Shelf - issues paper." Internal report for Monash University Library, Clayton, Vic. September 1996.
Monash University. Leading the way : the Monash Plan 1998-2002. [Clayton,Vic], Monash University, 1997.
Smith, Lisa. "Monash University Library Electronic Resources Directory: The fast track to Monash Library's electronic resources". Ariadne [online]. Issue 9, May 1997 [cited 1996-10-06]. Available from World Wide Web: <http://www.ariadne.ac.uk/issue9/electronic-resources/>. ISSN 1361-3200.