A digital library and archive about the Marshall Islands: experiences and challenges

Dirk HR Spennemann

Manuscript received December 2003

This is a refereed article

Introduction

The phenomenal growth of the world wide web since its popularisation in 1993 (Cailliau, 1995) has spawned a plethora of sites serving an incredible number of users. It is estimated that over 745 million people have internet connections, with the largest number (twenty-eight per cent) living in the United States (Global Reach, 2003). Many sites are government- or institutionally-sponsored and deliver information to the end-user. However, unlike the content of physical libraries, that of the websites is of varied quality, ranging from the authoritative to the mundane, often with limited veracity and authority, and including vanity publishing.

One of the major problems with web resources is the relative impermanence of many sites, resulting in the 'Error 404 File not found' message. While developed countries, such as Australia - through its Pandora project (NLA, 2003) - have been able to maintain archives of website content, smaller countries have not. As a result, much of what is published on the web is ephemeral in nature. At the beginning of the internet 'boom' it was feared that smaller countries might be electronically colonised by larger ones due to their lesser technological and economic base (Spennemann et al 1996). While the growing number of internet connections in smaller countries does not bear out this fact (cf. Global Reach, 2003), the electronic colonisation of the content has commenced. In a parallel development, the increasing commercialisation of the web has seen much of the amassed content being acquired, repackaged and locked up on commercial sites. This trend is bound to continue.

Structuring and accessing sources

Search engines such as Google and Altavista provide access to material, but return much without quality control. The web is anarchic in the delivery of information and disinformation and the power of the search engines means that only the top ten or twenty sites of the perhaps thousands of sites listed following a query are visited by users: further, the sites thus returned are of varied relevance, especially if the search term was too general. Early attempts at organising the material on the web were the Virtual Libraries (WWW VL, 2003a), such as VL-Pacific Studies (WWW VL, 2003b) and VL-Education (WWW VL, 2003c). Here links to various thematically related sites are moderated, assessed and the currency of the links maintained by competent individuals (WWW VL, 2003a). Compared to that centralised model, the Open Directory Project (Netscape, 2002) employs a large range of people with specific interests to compile a structured directory. Both virtual libraries and the Open Directory Project are subject to intellectual intervention and structured, resulting in some level of quality control. Nonetheless, many virtual libraries are in essence a collection of links to various resources distributed across the web, many of which in turn are collections of links.

The Marshall Islands

The Republic of the Marshall Islands (RMI) extends over about 1 950 000 km² of the Pacific Ocean but has only 171 km² of land. It comprises twenty-nine coral atolls and five islands in two chains trending north-west to north-east in an area between 4° - 19° North latitude and 160° - 175° East longitude. The environment is tropical, but, given the limited land mass, governed by oceanic climatic conditions, with limited vegetation coverage. While there is a large range of flora, the terrestrial fauna is limited.

European contact was essentially ephemeral and ship-based until 1859 when a permanent mission station and then trading stations were established on Ebon, a southern atoll. German trading interests gained ascendancy in the 1870s, leading to the formal annexation of the islands in 1885 and the establishment of a colonial administration from 1886 until 1914. Having taken them in the early stages of World War I, Japan was subsequently handed the Marshalls by the League of Nations for administration as Mandated Territory: she subsequently used the atolls as bases to launch attacks against United States possessions in Micronesia during World War II. The United States captured the islands in 1944 after a prolonged bombing campaign. Following World War II the Marshalls, like the rest of Micronesia, were administered by the United States as a Trust Territory on behalf of the United Nations. During the 1950s and 1960s the Marshalls, in particular Bikini Atoll, gained world-wide recognition as the location of nuclear and thermo-nuclear tests conducted by the United States. By 1989 the Marshalls had gained independence as the Republic of the Marshall Islands, closely allied to the United States through a 'Compact of Free Association' which, inter alia, permits Marshallese citizens to reside in the United States and large expatriate communities were established in Hawaii, Costa Mesa (California), Oregon and more recently in Arkansas.

Concept

The aims of the site 'The Marshall Islands. an electronic library & archive of primary sources' (http://marshall.csu.edu.au) are to:

1. provide a digital library and archive of material on the Marshall Islands,
2. make it available in an easy-to-use fashion,
3. design it to be usable efficiently with slightly out-of-date technologies and/or at slower connection speeds,
4. do so without fee as free-to-web publishing and without advertising, and
5. to ensure permanency of the material hosted.

The site started in 1999 with the publication of electronic versions of three hard-copy books: Marshallese legends and traditions (Downing et al 1993), Tattooing in the Marshall Islands (Spennemann 1992), and Essays on the Marshallese past (Spennemann 1993). Since then the site has been systematically expanded: it is hosted by Charles Sturt University (Australia), is and will remain non-commercial and free of advertising.

Table 1 Milestones in the development of the Marshall Islands site

Structure

The site comprises a total of 3452 text (html and pdf) and 7063 image files, structured in 230 directories. A further 300 text and over 1300 image files provide structural and navigational support as well as the 'look and feel' of the site. In total 1.63 gigabytes of data are uploaded on the server, with a further 0.53 gigabytes (in the form of over 900 text and over 2200 image files structured in 160 directories and sub-directories) in various stages of preparation.

Figure 1: Portal page of The Marshall Islands: an electronic library and archive of primary sources

All files mounted on the site are either accessed directly through search engines, or accessed via the site portal (Figure 1) and then through the structured first-level directory pages such as 'Culture and Society', 'Politics and Economy', 'European History', 'World War II', 'Environment', or 'Public Health', 'Literature & Art' and through a number of first-level directory pages that operate geographically, such as 'Map', 'Individual Atolls', and 'Country of Origin'. Most of the first-level directory pages have deeper-level overview pages, such as historic images or pages specific to major phases of Marshallese history. To provide a third approach, a list of all files hosted on the site can be accessed from the alphabetical list provided, as well as from a thematic file index which is in preparation. The portal page provides a 'link of the day' and a 'historic image of the day', both of which are generated from a table of calendar days, cycling at daily intervals within a monthly framework. These cater for the curiosity of 'accidental tourists' arriving at the site. In addition a 'What's New?' page and a page of Editorial Comment has been provided.

As with any structuring of content, there are problems of classification, with some material belonging to more than one possible category. Unlike physical libraries, where volumes can only be shelved at one location, digital libraries allow for multiple entities and thus alternative pathways of access. The site is structured through easily recognisable classes, which are all represented at the portal page level. There are two exceptions to the over-riding logic: the categories of 'World War II' and 'Stamps.' Both were placed on the portal page after extended correspondence with other website managers, on the ground that these cater for high-volume special interest groups. The usage statistics certainly bear this out.

Figure 2: Structure of the first and second-level directories of 'The Marshall Islands: an electronic library & archive of primary sources'

Page design

The emphasis is not on form, but on the quality of the data presented and on the permanency of the data sources. Consequently, the main directory pages as well as those for the subordinate level of directories are designed to be image-poor, so that they load faster and enable users with older, and thus slower, machines to access the material and/or connect via slower modes of access, as is generally the case in developing nations. Document pages may contain a number of images in which case limited resolution copies are mounted on the page, with each linking to a higher quality image presented on a separate page.

The overall lay-out is elegant, yet plain, avoiding gimmicks, animated images or graphic front pages, which swiftly show their age. The text is black on white background, using serif fonts (Times) for body text and sans serif fonts (Helvetica/Arial) for headings. All pages have a similarly-designed masthead with an image that varies between the major categories. Likewise, all pages have a footer which provides site navigation options as well as bibliographical information. The description of webpages has been problematic for librarians developing citation and style manuals, and an abundance of possible approaches exists: to influence the outcome a preferred form citation is offered at the bottom of each page, thus:

Spennemann, Dirk HR (2002). Postage Stamps used in the German Marshall Islands. German Postal Services on Jaluit. http://marshall.csu.edu.au/html/Stamps/Stamps_History_Jaluit.html

Since early 2003 all pages have been designed with a fixed text width of 640 pixels, which prevents users from widening or narrowing the text on the screen. This has been done intentionally to ensure maximum readability by limiting the body text to about 60 characters per line which is deemed to be the ideal line length, and allows a white 'margin' at the left and right which adds readability (Youngman & Scharff, 1998). Furthermore, the fixed width of the text block allows us to 'format' documents so that images remain at their intended placement. Progressively, older pages will be reformatted.

Language

The language used for the pages and documents is English. While some German archival documents are reproduced on the site, where possible English translations are provided side-by-side. Two bilingual children's books exist in English and Marshallese. Clearly, the language of the user limits access to the library. Other research has shown that non-English speakers prefer to read websites in their own language rather than English (IDC 2001 quoted after Global Reach, 2003), with thirty-four per cent of French users being prepared to read English-language pages, compared to only eighteen per cent of German users, and as little as eight per cent of Japanese users. Global Reach (2003b) also suggests that at present some 35.6 per cent of the world's on-line population speak English. Projections suggest that the number of non-English speaking on-line users is likely to increase dramatically: realistically, it is not possible to make too many of the documents on the server available in languages other than English, with the possible exception of narratives to be developed for major subject areas.

Usage

Rather than counting 'hits', system calls to the server which include calls to image files, the statistics presented here only account for html and pdf files. Some web browsers, such as Internet Explorer re-deliver a cached page image without downloading again. Thus the actual number of 'viewings' of a given page is higher than the number of requests sent to the server: this is particularly true for directory and overview pages. As a result the server log statistics only count ' pages delivered' or 'impressions.'

How many impressions?

The statistics for the site have been collected by three different means, which causes some problems of comparison. The first counter was activated on 17 February 2000, but only measured the frequency with which the portal page was called up. Subsequent counters were placed on second-level pages such as culture, history and politics. As most search engines index and then target individual pages, and because many users may never access the site portal, these early statistics grossly undercounted total usage. The statistics until March 2002 were kept courtesy of Netscape's web.hitometer.com On 6 September 2001, a new counter was established to measure the total usage of all pages on the site. That counter was set to commence counting at 44 948, which was the sum of the usage of all main pages with counters to that date. This underestimates the total site usage, but was the only verifiable value available. From 6 September 2001 on, all sub-pages had the same counter so that direct visits to lower-order pages contributed to a unified overall site visit count. A comparison between the total site count and the count of the visits to the portal page alone (period October 2001 to August 2002) suggests that less than ten per cent of the visitors use the portal page at any point of their visit. Based on this ratio, the total site count on 6 September 2001 would have been closer to 260 000 than the 44 948 used as the starting point of the current figure.

Commencing April 2002 accurate server-level statistics (through logs) became available. Combined with the fact that the hitometer was discontinued in August 2002, the counter system had to be revised. Some of the statistics provided by the hitometer service can be compared to the data from the server logs. A comparison of the count of pages delivered for the period 21 March to 30 June 2002 showed that the hitometer statistics undercounted the actual delivery of pages by an average of 16.2 per cent due mainly to the fact that those users who do not allow 'cookies' to be placed were not counted by Netscape's hitometer system.

In May 2003 the site was moved to a new server. After this move, the number of log entries dropped by seventeen per cent because the new server has a different page request log reporting system. Figure 3 graphs the development of the adjusted statistics, with the post-May 2003 data plotted uncorrected. During October and November 2002 the current maximum monthly usage rate occurred, with over 120 000 pages delivered: in the second part of 2002 and early 2003 the monthly average was around 85 000, and since then it has declined to just over 70 000.

Figure 3: Pages deliversed between February 2000 and November 2003

How many users?

Given that a single user can request a number of pages in succession, the total number of impressions provides a useful statistic for the relevance of the digital library, but does not provide an insight into the number of actual users. The server log files, however, allow the assessment of the number of unique hosts requesting page impressions. This is equivalent to the number of unique users. Except for situations where a user is dynamically allocated an IP address, as in the case of some dial-up connections, these unique hosts actually represent unique users, irrespective of whether they came and went on visits several days apart. The frequency of unique users as shown in Figure 4 demonstrates an average in excess of 12 500 unique users for the past twelve months. To place this in perspective: unique users of the Marshall Islands site are equivalent of about ten per cent of all unique users accessing the entire public web server of Charles Sturt University, and about eight to nine per cent of all unique users accessing that of the University of Melbourne.

Figure 4: Number of unique users between March 2002 and November 2003

Geographic origin of users

In total, there are 247 extant and historic, country domains (ISO, 2003), 156 of which, or 63.1 per cent, are represented in the logs for the Marshall Islands site since its inception (Table 2); see Table 8 for details for 2002 and 2003 records. These are illustrated in Figure 5, showing that with the exception of a few countries in Asia and South America the major area without requests is Africa. While total number of users is very low there, all African countries have internet access at least in their capital cities (APC, 2003).

The volume of page impressions requested can be compared with the total number of internet users per country as estimated by Global Reach (2003). Users from United States domains dominate the requests, but it also has the highest number of internet users. The figures are roughly comparable until we include page requests from .com domains (Table 3) as well. Then the United States dominate totally. Apart from the United States the greatest discrepancy can be observed for Australia. While it makes up only 1.7 per cent of web users, between 17.9 and 21.9 per cent of page requests originated here. Both the case for the total United States as well as for Australia underline, of course, a salient fact: demand for information is not uniform, but driven by the needs of the clientele. The United States has historic and current political ties with the Marshall Islands, while the Pacific is, current political rhetoric notwithstanding, essentially Australia's backyard. Lack of interest in the subject matter is understandable due to political and historic factors when we consider the relative under-representation of China, the Russian Federation or India. Yet, geographic distance is not necessarily a deciding factor, as the examples of Germany (under-represented by fifty-five per cent) and the Netherlands (over-represented by 270 per cent) demonstrate. Despite the Marshalls once being a German colony, it seems that linguistic ability (more Dutch than Germans are fluent in English) and overall intellectual outlook (Germany on the whole currently being more inward looking) are influencing the use of a digital library.

Some requests are understandable in view of the geographic similarities of the requesting nations, like small island states and countries, such as Malta, the Faeroe Islands and several small states in the Caribbean. But why, then, are there requests from some of the poorest countries in Africa without connections to the Marshall Islands, and without geographic similarities, such as Ivory Coast, Ghana, Benin and the like? The answer lies perhaps in the incessant stream of the digital version of the infamous Nigerian mail scam and its variations. It is very likely that at least some these requests are part of an e-mail harvesting program.

(*) combined total of .edu /.gov /.mil/ and .us domains.
Table 2: Geographic origin of users: top fifteen domains (based on 2003)

Figure 5: Geographic origin of users (dark areas and dots)

In addition there are users from the domains of business (.biz), commercial (.com), network (.net), organisation (.org), international (.int) and the like (Table 3). These are usually United States addresses, but given the ubiquity of United States companies in the global telecommunications market place, some may well be the sole providers for entire countries (cf palaunet.com). A substantial percentage of user origins cannot be resolved at all, as the servers are only identified by their numeric IP address. While it is possible to identify the location of each server, it is not economical to do so.

Table 3: Geographic non-specific user domains

Some IP addresses of commercial domains (.com .net) allow further breaking down, depending on how the service providers code their regional server hubs. For the major United States and Australian providers this allows a breaking down to at least the state level which is sufficient for the purposes of understanding the origin of users. Table 4 sets out the regional breakdown for the United States for the periods July to December 2002 and for 2003 (excluding December), while Table 5 provides the regional breakdown for Australia. For the United States the 2003 pattern reveals that populous states, such as New York and California, together with North Carolina and Hawaii are heading the list. The high percentage of Hawaiian users is understandable as the island state has connections to the Marshalls and the Pacific in general. The high percentage from North Carolina cannot be explained (Table 4). The Australian data (Table 5) reflect population densities across the continent as well as the fact that the eastern seaboard borders the Pacific.

Table 4: Origin of page requests from US commercial providers

Table 5: Origin of page requests from Australian commercial providers (Jan to Nov 2003)

Another measure of the geographical distribution of the users across the globe is the time of demand on the server to deliver pages. At any given location, the demand is likely to be higher during the waking hours of the user, peaking during office hours, while troughing out in the early hours of the morning, reflecting a largely diurnal curve. The hourly demand curves for individual locations would cancel each other out were the users evenly distributed across the globe, resulting in a perfectly flat line. In the light of the vast expanse of the Pacific Ocean, this is unlikely to occur. However, as a rule, the flatter the curve, the more geographically even the demand. Figure 6 demonstrates that while in 2001 the demand reflects the diurnal pattern it has become increasingly global since then. Given the overall large percentage of United States-based users (Table 2) it is not surprising that the curve maintains a weak diurnal pattern.

Figure 6: Average hourly demand for pages (in per cent of requests per day) for 2000 to 2003, compared to the ideal load (time zone is Greenwich Mean Time)

What users are looking for

Any library will provide for a wide range in clientele. We are no different in that regard. Table 6 lists the number of pages, both html and pdf, mounted on the site in November 2003 and compares this with the average monthly requests for pages. The pages have been grouped into major subject areas. A number of pages provide individual historical pictures and their sources. These have been assessed separately. The number of pages by subject area is not necessarily indicative of the relative significance of this area to the manager of the website, but may be a function of the way the information is presented. The nature of postal history and philately, for example, required that the information be broken up more than other subject matter, resulting in a greater number of pages. Table 7 allows for this and provides the average frequency with which each page in the subject area was called up per month.

(*) Historic photographs of both traditional cultural subjects and European Colonial History, excluding World War II.

Table 6: Average number of pages delivered per month (by subject area) between April 2002 and November 2003 in comparison with actual pages mounted on the server

Table 7. Average frequency with which each page in a category was requested per month

The adjusted monthly averages show that pages dealing with Word War II as subject matter had the highest request rate with each image page being requested on average 142 times and each text page sixty-three times. Given anecdotal evidence of high levels of activity by WWII buffs on the web this was not surprising. What was surprising, however, is the high impression rate for bibliographies (67.4). At the other, low end of the scale, as expected, are administrative pages (editorial, what's new, file lists, link of the day), but surprisingly also historic images.

The statistics show changes over time, with demand on some subject areas, such as the environment, in steady decline and other areas, such as bibliographies and literature and art, on a steady increase. These data allow the editor to respond to demand and update the relevant pages more frequently, but also give cause to rethink the approaches in less frequently requested subject areas.

Table 8: Geographic origin of users of the Marshall Islands site compared to the estimated population of internet users in that country

Table 9: Ranking of the site 'The Marshall Islands' in various popular search engines

Status and visibility

Before we look at the challenges faced in maintaining such a library, some comments are in order to assess the relative status and visibility of the site by assessing how search engines, the main access tool to information by the average user, rank it.

A search for the generic term 'Marshall Islands' was run on 17 December 2003 using Google, the most popular engine at the present time: it found approximately 3 610 000 pages with the term 'Marshall Islands' somewhere on the page. The site discussed here ranked fourth after the official page of the Republic of the Marshall Islands' Embassy in Washington (which because of its official status will always rank first, regardless of breadth or depth of the site itself); the CIA World Factbook which ranks highly for all smaller countries, again because of its perceived official status; and the official site running the Marshall Islands internet name register which contains no information on the islands at all. Table 9 provides a comparison of the top ten Google rankings with those of six other popular search engines. In all cases the Republic of the Marshall Islands' Embassy site was ranked number one, while the sites under discussion in this paper ranked in the top five.

Challenges

One of the challenges is to deal with the number of information requests that flow in. In order to allow users of the web pages the capacity to ask questions, an e-mail address was provided at the bottom of each page from the start. An unwanted side effect of a web presence is that websites are being trawled by web crawlers and robots for e-mail addresses which are then sold on to marketers sending out mass e-mails. As a result of the inclusion of the e-mail facility on the bottom of pages, SPAM has increased disproportionately (approximately 30-fold) to that of fellow academics working in the same school. With the aid of filters, it was decided that this is a 'price' worth paying. Some of the legitimate questions are those faced by any librarian: requests for help expressed at various levels, from professionals requesting detailed information after other avenues have been exhausted, to high school and undergraduate students requesting that someone writes their essay for them. These are responded to differentially, given my limited resources and the other demands on any academic's time.

The site also acts as a magnet for other material being 'donated.' While it has long been known that sites attract visitors and direct them to other material, the 'honey-pot effect' (Green 1995), demonstrated here is an unexpected variation to the theme: private archival material has been offered for inclusion in the site, which adds to its richness, and enhances the research capacity of the site's editor.

Conclusion

There are a number of ways of accessing information on the web: while the initial fear of electronic colonisation of smaller countries may be unfounded, the locking up of content through increasing commercialisation may exclude many economically weak participants from accessing it. There remains a need for 'free-to-web' libraries of primary and secondary source materials rather than mere links to other collections.

On-line server logs allow study of the library users and the subject requests made by them. If a digital library is to be relevant to the audiences it presumes to address, in-depth analysis is a necessity: it can suggest how the library 'collections' can be added to and restructured to make content with apparent low demand more accessible to potential users.

This case study of a digital library focussing on the Marshall Islands suggests that there is world-wide demand for such content: given that amongst the material much is of limited distribution in paper form, or is transcribed from rare and inaccessible archival resources, it provides an essential service in the dissemination of information.

Acknowledgements

The site opened in 1999 with the publication of three electronic books (Marshallese legends and traditions, Tattooing in the Marshall Islands, and Essays on the Marshallese past), the raw coding for which was set as class exercises at Charles Sturt for the subject 'ITC 130 On-Line Publishing' with the co-operation of Professor David Green, Rob Stocker [and the students]. Darren Stuart provided some of the further site development and re-coding of the students' work on the three books into the present form. Since then the site has been systematically enlarged and enhanced by the editor, in which I am particularly indebted to all authors who graciously made their material available, as well as to the web administrators who shared lists and data collected by them.

References

APC (2003) The Association for Progressive Communications. Information & Communication Technologies (ICTs) Telecommunications, internet and computer infrastructure in Africa. http://web3.sn.apc.org/

Cailliau, Robert (1995) A little history of the world wide web from 1945 to 1995. World Wide Web Consortium. http://web.w3.org/History.html

Downing, Jane, Dirk HR Spennemann and Margaret Bennett (eds) (1992) Bwebwenatoon Etto: a collection of Marshallese legends and traditions. With illustrations by Amram Enos and Paul Kingsbury. Marshall Islands Culture And History. Series E: Legends, Chants and Proverbs, vol 1. Majuro, Republic of the Marshall Islands: Historic Preservation Office.

Global Reach (2003a) Details of country/language analysis. Last revised on 30 September 2003. http://global-reach.biz/globstats/details.html

Global Reach (2003b) Global internet statistics (by language). Last revised on 30 September 2003. http://global-reach.biz/globstats/index.php3

Global Reach (2003c) International online marketing to increase your web traffic - and sales - from abroad http://web.global-reach.biz/eng/index.php3

Green, DG (1995). 'From honeypots to a web of SIN - building the world-wide information system'. In: P Tsang, J Weckert, J Harris, and S Tse (eds) Proceedings of AUUG'95 and Asia-Pacific World Wide Web '95 Conference, Charles Sturt University, Wagga Wagga, pp11-18.

ISO (2003) International Standards Organisation. VISO 3166 Code lists. English country names and code elements. http://web.iso.org/iso/en/prods-services/iso3166ma/02iso-3166-code-lists/list-en1.html

Netscape (2002). About the Open Directory Project. http://dmoz.org/about.html

NLA (2003) National Library of Australia. Pandora Archive - preserving and accessing networked documentary resources of Australia. http://pandora.nla.gov.au/index.html

Spennemann, Dirk HR (1992) Marshallese tattoos. Marshall Islands Culture and History. Series C: Traditional Skills and Practices, vol 1. Majuro Atoll, Republic of the Marshall Islands: Historic Preservation Office. 233pp.

Spennemann, Dirk HR (1993) Ennaanin Etto: a collection of essays on the Marshallese past. Marshall Islands Culture and History. Series F: Technical Studies and Miscellaneous, vol 1. Majuro Atoll, Republic of the Marshall Islands: Historic Preservation Office. 314pp.

Spennemann, Dirk HR Jim Birckhead, David G Green and John S Atkinson (1996) 'The electronic colonisation of the Pacific.' Computer Mediated Communications Magazine vol 3, nº 2. http://web.december.com/cmc/mag/1996/feb/spen.html

Williams, Stayce and Lauren Scharff ( 1998) 'The effects of font size and line spacing on readability of computer displays'. http://hubel.sfasu.edu/research/SWExp.html

WWW VL (2003a) 'The WWW Virtual Library'. http://vlib.org/

WWW VL (2003b) 'Pacific Studies' WWW Virtual Library http://coombs.anu.edu.au/WWWVL-PacificStudies.html.

WWW (2003c) 'Education Virtual Library' The WWW Virtual Library. http://csu.edu.au/education/library.html

Youngman, Melissa and Lauren Scharff (1998) 'Text width and margin width influences on readability of GUIs' (Presented at SWPA 1998) http://hubel.sfasu.edu/research/textmargin.html

Dirk HR Spennemann can be reached at the Johnstone Centre and School of Environmental and Information Sciences, Charles Sturt University, PO Box 789, Albury 2640, dspennemann@csu.edu.au

Biographical information

Dirk HR Spennemann can be reached at the Johnstone Centre and School of Environmentals and Information Sciences, Charles Sturt University, PO Box 789, Albury 2640, dspennemann@csu.edu.au.