Fijian Spell Checker for OpenOffice

1) Coordinators

Dr. William J. Blanke (primary)

Dr. Paul Geraghty                            

2) Project Objectives

To create a spell check dictionary for the Fijian language to be used with the OpenOffice free/open source software suite of office programs.

3) Expected Outcomes

OpenOffice contains a mature, feature rich word processing application that is compatible with Microsoft Word’s file format. Adding a Fijian spell checking dictionary to this free/open source software should strengthen the use of the Fijian language in print.

4) Beneficiaries

• Any organization such as newspapers, religious groups, governments, or educational institutions that write documents in Fijian
• Students who are learning the Fijian language
• The Fijian language itself, through increase in quantity and quality of printed material

5) Project Description

OpenOffice is perhaps the most popular free/open source software word processing program in existence and is available on Linux, a free/open source software operating system, as well as many other platforms. Its spell checker has support for over 37 languages. Aside from Maori, languages from the Pacific islands are lacking.

This project will create a Fijian spell check dictionary for OpenOffice by taking an existing Fijian dictionary as provided by Dr. Paul Geraghty from the University of the South Pacific (USP). This dictionary is not in the correct format, so a computer program will need to be written by USP computing sciences students in order to create a word list from this vast original text. Additionally, students familiar with Fijian will need to add grammatical rules so that OpenOffice can process suffixes and prefixes properly. Eventually, more students familiar in Fijian will need to test the final product by spell checking existing Fijian documents. Computing sciences students will need to provide technical assistance throughout the project.

5.1) Background to Project

USP has some of the best Fijian language scholars in the world. It also has a very competent computing sciences department with students talented in programming. Because of this, it is suited to produce a spell check dictionary that is accurate and thorough. In addition, the students at USP are another great resource as many are native speakers of the Fijian language.

Dr. Paul Geraghty has worked on the Fijian dictionary project and is able to provide the latest draft of the iVolavosa Vakaviti (Monolingual Fijian Dictionary). This text will form the basis for the Fijian spell check dictionary. However, before it can be used, it must be manipulated into the correct format for OpenOffice to understand. That role will be filled by students in the computing sciences department. Using free/open source software tools such as the gcc and g++ compiler on the Linux operating system, they will create a program to sift through the text of the iVolavosa Vakaviti to generate an OpenOffice compatible dictionary.

Throughout the entire process, native Fijian speakers in the USP student body will be able to proof the progress of the project—identifying any errors that could appear. This should greatly increase the accuracy of the spell checker. These students can also assist in creating the logic OpenOffice uses to parse variations of words. Fijian doesn't have plurals, but it does have regular derivations that are not all covered in the dictionary, particularly with prefixes vaka- and va- and suffixes such as -taka. Older spelling conventions would also have to be considered.

5.2) Justification for Project

Computers and the Internet mean more and more text is being dealt with in electronic form. Electronic documents are easy to transmit and store, and having a document in electronic format allows the use of advanced word processing tools that would not otherwise be available. The most widely used tool is the lowly spell checker, which lets even the worst speller appear flawless and is the best first defense against typographical errors.

To our knowledge, no Fijian spell check dictionary currently exists. This means that any document written in Fijian must be manually proofread. For large documents, this can be a daunting task. Therefore, the use of Fijian in developing correct documents is more difficult than in other languages.

This project should eliminate that obstacle. Hopefully it will encourage more use of the Fijian language in print and on the Internet. It should also serve as a solid reference for proper spelling of the language. OpenOffice and its spell checker are free/open source software and can run on Linux, a free/open source software operating system. Therefore this complete set of tools will be free for anyone in the community and thus should aid in the sustainable development of Fiji.

6) Deliverables

On August 11, 2005, the project was completed and the following files were submitted to OpenOffice.

README_fj_FJ.txt

fj_FJ.dic

fj_FJ.aff