Curriculum Vitæ

• 1994-1998: PhD in Linguistics & Applied Linguistics. (Supervisor: Dr Jean Mulder) (Awarded 1999-10-29)
• 1993-1994: Master of Engineering Science (Cognitive Science; Masters by Coursework and Shorter Thesis) (Supervisor: Dr Lesley Stirling)
• 1989-1993: Bachelor of Science (Computer Science)
• 1989-1992: Bachelor of Electrical Engineering (with Honours)

• 1985-1988; Victorian Certificate of Education (HSC)

The FRED (Federated Repositories in Education) project involves developing infrastructure for repository federation in the education sector in Australia. My responsibilities included:

• establishing user requirements through use scenarios and use cases, gathered through conversations with stakeholders and review of literature
• developing a reference model and a service usage model for repository federation
• specifying individual core services for repository federation, at the abstract and concrete service levels, through the e-Framework approach
• presenting project results and tutorials in workshops and seminars

The PILIN (Persistent Identifier Linking INfrastructure) project involves prototyping and scoping infrastructure for Persistent Identifiers in Australia, with a Service-Oriented Approach to implementation. My responsibilities included:

• establishing user requirements through use scenarios and use cases, gathered through conversations with stakeholders and review of literature
• developing a reference model and ontology for identifiers, identifier services and identifier systems
• writing policies and guidelines for identifier use, outlining options for governance, architecture and implementation
• specifying service usage models for identifier management systems, through the e-Framework approach
• presenting project results in workshops and seminars

I co-supervised a Masters thesis in sociolingustics, on the manipulations of language-based constructs of identity taking place in the Valle d'Aosta, a multilingual autonomous region of Italy. The dissertation makes eclectic use of Social Identity Theory, Ethnolinguistic Identity Theory and the notions of diglossia and language usage domains, in order to model competing profiles of linguistic identity construction in the region. It is based on both wide-ranging questionnaires, and interviews with members of the political elite of the valley, as opinion leaders.

I was responsible for the IT needs of the department at large. This has included:

• desktop support (on Mac and PC), both over the phone and in person;
• incidental project work and software support specific to language teaching and research needs;
• training staff in use of software and audiovisual equipment, both one-on-one and in groups;
• troubleshooting of software, system and hardware faults (Mac, PC, A/V);
• administration of filesharing server (Apple) and backup server (Retrospect);
• administration of local web server, with online database access, scripting, bulletin board, online survey sites;
• advice, budgetting, procurement and deployment of hardware and software purchases;
• maintaining departmental IT assets inventory.

Providing research, editorial, programming, and administrative assistance to A/Prof John Hajek in his research work. This includes:

• proofreading & editorial work;
• linguistic analysis of data (phonological, syntactic, historical);
• data analysis and extraction using computer tools (scripting, databases);
• bibliographical research;
• development and maintenance of computer databases of linguistic data.

From 1999 to 2001, I was employed at the Thesaurus Linguae Graecae (TLG) project at the University of California, Irvine. The TLG is a digital library of Classical and Byzantine Greek literary and sub-literary texts. I have generally been involved with issues surrounding the text encoding employed in the project. Specifically,

• From February to April 1999, and intermittently since, I have carried out most of the development of the in-house Verification and Correction tools, Argos and Io. These have been developed in Visual Basic on Windows NT. Argos performs format checking: it converts the texts from the format received from data entry (Alpha code) to that used within the project (Beta code), tallies the non-alphabetic codes used in the texts for checking, and highlights potential problems in the texts (erroneous citations due to line splits, mismatched brackets and quotation marks, misplaced punctuation). The program also incorporates text search functions, and an RTF rendering engine of Greek text to allow browsing. Io incorporates an accentuation checker, verifying that the accentuation of words follows Classical Greek norms, and a spell-checker, working on word lookup of canonical forms of the words from a database.
• From May to November 1999, I worked on producing TLG CD ROM #E. The major component of this work (May-July) involved reverse-engineering the existing CD ROM #D, whose binary format was inadequately documented, in order to allow backward compatibility with existing TLG CD ROM-compatible software. This included determining the criteria whereby texts were to be subdivided into sections in the electronic tables of content. The other significant component (August-September) was deriving a word index of the entire corpus. This involved not only normalising word forms according to standard Greek convention, but also determining complete and fragmentary word termini in the involved typographical conventions of papyrological and codicological editorial practice. Software was written in C on Windows NT.
• From December 1999 to June 2000, with intermittent maintenance since, I was involved in developing a search engine for the TLG corpus on the web. The search engine includes both word index lookup (connecting to a MySQL database of word forms and a B-Tree index of word instances), and textual search. The latter includes both literal text search and regular expressions; it also allows for the search of orthographically normalised strings in the text (e.g. ignoring secondary accents), and optionally ignores diacritics and case. Text searches can also skip over stretches of non-text interrupting words (including symbols and annotations). The idiosyncrasies of the encoding of Greek made it necessary to develop efficient finite state automata customised to these requirements. Searches may also involve context, specified in either words or lines. The search engine also incorporates a variety of display modes, depending on user-specified font selection and desired detail of rendering. The software also communicates with the bibliographical component of the search engine, stored on a Postgres database. Software was developed in C on Windows NT and Solaris. Also authored user documentation.
• In December 1999, and July to August 2000, I documented the usage in the project's corpus of Beta code, the encoding scheme whereby the Greek text and non-textual entities are encoded. This involved some 800 entities, many of which had been used inconsistently over the years, and most of which had been poorly documented to date. This necessitated scanning the corpus exhaustively, and repeatedly consulting the physical editions from which the electronic texts had been derived. As a result, I developed guidelines for text coding, and corrected many instances of misuse of Beta Code, involving both confusion of entities and the correction of past miscoding of text; these resulted in substantial improvement of the quality of the corpus' word index. This work contributed to the proposals for new characters submitted to the Unicode Consortium by the TLG; and I have continued to consult the Unicode Consortium on issues of character encoding specific to Greek (including archaic alphabetic variants and Byzantine Musical notation).
• From September 2000 to October 2001, I worked on the conversion of the corpus from Beta code to TEI-conformant XML. Since the TLG has consistently followed a policy of reproducing the printed page literally in its electronic text, the conversion process is involved, and subject to the editorial and typographical idiosyncracies of a wide variety of texts (close to 7000 in number). The conversion took place in Perl. I also expanded the existing XSLT stylesheets for the TEI tagset to deal with the breadth of features present in the corpus.
• Since May 2003, I have been working on lemmatising the TLG corpus; this requires a good familiarity with all stages of Greek represented in the corpus, from Homeric Greek to Modern Greek dialect. The project has involved extensive linguistic research and programming of rules and phonological processes, to deal with the wide range of linguistic forms represented, as well as extensive testing and incremental addition of stems and inflections.

I have been lecturer for the subjects Introduction to CALL Project, a course involving teaching HTML through Dreamweaver and Web Design principles, as applied to a Computer-Aided Language Learning project. This has involved devising course content, lecturing, technical assistance with web publishing, and assessing and managing project work. In 2006, I also took three lectures in the Introduction to CALL subject proper, familiarising students with the range of technologies available for CALL, as well as the challenges in fitting the technologies to the paedagogical needs rather than allowing the reverse.

I have been course coordinator, lecturer, and tutor for the subjects Introduction to Language (First Year; first semester, 2002 & 2003) and Historical Linguistics (Second/Third Year; second semester 2002). This has involved developing course content (for th most part from scratch), web publishing course content, compiling course handbooks, lecturing, pastoral work, devising and assessment assessment, and coordinating tutors.

I was employed to realise two projects:

1. Skryba, a spelling practice program on the Web (using Javascript and Perl). This required a knowledge of orthographic conventions and phonological rules, used in a gapping exercise where users type in missing letters in orthographically complicated words. The current project involves Russian, although the program is intended to be language-independent.
2. Programming and publishing on-line course materials for the Intermediate Swedish course of the University. This involves HTML formatting, Perl scripting, cookie-based authentication management, and database administration (MySQL), on a MacOSX platform.

For the Logical Language Group (a non-profit educational organisation), I have been involved in editing and publishing on-line an introductory brochure and textbook. While the content was developed collaboratively, I have been solely responsible for the technical aspects of publishing the materials; this was done in XML (Docbook) and DSSSL, working with Jade, and generating RTF, TeX and HTML output. I have also extensively customised the Docbook DSSSL stylesheets, and offered bug fixes and improvement suggestions to the standard Docbook DSSSL stylesheet suite.

I worked on the University Computer Helpdesk handling queries from throughout the campus on both Macintosh and Windows platforms, providing desktop, lab, and hardware support. I was also involved in writing documentation and helpfiles.

While a Masters and doctoral student in the department, I participated in various research programmes as a programmer:

• From October 1996 to February 1997, I was a research assistant for Dr Dominique Estival. Dr Estival was carrying out a project involving analysis of syntactically-tagged corpora of Old French. My involvement was in converting an electronic text (the Vie de Saint Louis by Joinville) from a Microsoft Word file to SGML-coded text (TEI Lite), and setting up the infrastructure for syntactic tagging and parsing of the texts, including obtaining and customising SGML software. The conversion was undertaken mainly in C; the parser was in TCL.
• From December 1993 to June 1995, I was involved in developing computer software for Dr Mark Durie, University of Melbourne, used to determine the correlation between the referring status of nominal expressions in a text, and their linguistic features. The software was based on the relational database 4th Dimension on the Macintosh, and involved extensive scripting to enable efficient coding of the texts, as well as statistical computations based on the aggregate relative distances between the coded entities in the texts.
• From September 1993 to January 1999, I was a research assistant for Dr John Hajek, University of Melbourne. My main involvement was with his postdoctoral research---a survey of the phonetic and phonological systems of the languages of the Pacific area. I was involved in database software development, data analysis, and data entry. The relational database used was 4th Dimension on the Macintosh, and the software included automatic decoding of IPA characters into phonetic features, and plotting data points on zoomable geographical maps.

I presented courses on the software packages Microsoft Word, Endnote, and Adobe Photoshop, in both Windows and Macintosh, and an introductory course on the World Wide Web.

I was a computer laboratory demonstrator and tutor for the departmental first year programming courses on the functional language Miranda (Computer Science) and on Pascal (Computer Science, Statistics), on the Macintosh. The Statistics course also included some instruction in elementary statistics.

In January and February 1995, I held an internship with the Language Technologies Group of the Microsoft Software Institute, Australia, working under Dr Robert Dale. There, I developed a rudimentary text planner and generator producing encyclopaedic texts on animals from database information. The text generator, PELUDO, was since continued in Maria Milosavljevic's doctoral work on PEBA II. The text planner was template-based, and programmed in Prolog on Windows NT.

Mathematical programming simulating the effectiveness of mine neutralisation, in Turbo Pascal on DOS.

Setting up a relational database for the library holdings of the firm, in Paradox on DOS.