Internet Archaeol 17. Falkingham.

Section 4: A Practical Evaluation of XML Technologies and TEI P4 for Archaeological Markup and Multi-layered Presentation

4.5 Evaluation and Conclusions

In relation to the aims identified in 4.2.1 above, the encoding of the electronic reports has successfully created valid, well-formed XML documents that can be used indefinitely to create an almost exact replication of the original document. Structural and thematic encoding has been successfully applied using the XML version of the TEI P4 Guidelines (Sperberg-McQueen and Burnard eds 2002). A number of basic XSL stylesheets have been written to provide a range of effective document transformations designed to meet a variety of user needs.

The encoding of archaeological content has, to a large extent, been achieved by the application of attributes to elements, and referencing of external thesaurus terms, as the full range of TEI P4 elements alone was not extensive enough to describe archaeologically-specific subjects. This is not surprising, as the Guidelines have been created primarily for use with, and analysis of, literary and linguistic texts. However, as a 'proof of concept' practical case study, the encoding of the digital grey literature reports and their client-side transformations have demonstrated how XML technologies and TEI P4 can be used together effectively.

Tailoring the display of the reports for varying user needs, based on both structure and content, can be done with relative ease by the application of basic stylesheets. XSLT can be seen to be a far superior method of styling and transforming content than use of CSS. Client-side processing was chosen due the individual circumstances of the author and there have, as a result, been limitations in what may be achieved due to lack of Web browser support, primarily with regard to CSS styling (see 4.4.2). It is acknowledged that server-side processing is to be preferred, although additional hard- and software resources would be required to implement this.

The author's manual application of the XML encoding using Microsoft Notepad was a time-consuming process that could have been accelerated by the use of an XML text editor (see 3.3.2). Becoming familiar with the full depth of the XML version of the TEI Guidelines also took a considerable time in order to identify the relevant elements and attributes to use; nevertheless, TEI-conformant documents have been produced, which include the range of desired encoding identified following document analysis. It must be noted, however, that the encoding has been applied according to the author's own interpretation of the Guidelines, particularly the TEI Header, the rules for which, as Morrison et al. (2002) acknowledge, are flexible. Any future archaeological application of the Header, and indeed document markup as a whole, would need to agree upon a standardised approach. Virtually any part of a text may be encoded to increase the potential and the power of document transformation. However, such additional input of resources needs to be balanced against both the immediate and longer-term gains of doing so.

The lack of certain data within report content was a limiting factor, particularly in retrieving some OASIS Project-related output, in particular information regarding archival location and content, and the attribution of dates to particular features (see 4.4.3.2). Image quality was also an issue, having worked with copies of original illustrative material, yet the ability to retrieve images for Web display through the use of XSLT was a satisfying achievement.

The author came to this project as a novice in the use of XML and XSLT, whose knowledge increased greatly as the case study progressed. Whilst it has been possible to achieve the desired aims, it is acknowledged that this is just a starting point; more advanced users of XSLT will, no doubt, be able to produce more sophisticated ways of applying styling and transformation.

Nevertheless, it is hoped that this exploration of the potential of archaeological document markup has developed the subject beyond those of previous exponents of XML (Gray and Walford 1999; Meckseper 2001). The multi-layered presentations have demonstrated how XML and XSLT have the power and flexibility to open up new possibilities for the presentation of grey literature on the Web above and beyond those achievable with the proprietary file formats favoured at present. There is national interest in, and call for, the development of new methods of electronic publication for archaeological reports; it is hoped that this dissertation will contribute to this debate (Williams 2003; Miles 2004). A number of humanities projects have already begun to experiment with and exploit the potential of XML in the provision and manipulation of textual content online; it is hoped that as such use becomes more widespread, XML will gain increased acceptance and support within the discipline of archaeology.