SEVENTH AUSTRALIAN CONFERENCE ON NEURAL NETWORKS (ACNN'96) PRE-CONFERENCE WORKSHOP ON 'ACTUAL AND PROSPECTIVE USES OF NEURAL NETWORK TECHNOLOGY AS AN AID FOR FORMS PROCESSING IN THE AUSTRALIAN PUBLIC SERVICE (APS) This page was last updated on: 4 April 2019 On Tuesday, 9 April 1996, a pre-conference workshop on 'Actual and prospective uses of neural network technology as an aid for forms processing in the Australian Public Service' was held as a joint ACNN'96/ACS/OTSIG/RMAA initiative on the afternoon prior to the commencement of the Seventh Australian Conference on Neural Networks (ACNN'96 - to be held from 10 to 12 April 1996, at the Australian National University, Canberra, Australia). The venue for this workshop was the Huxley Lecture Theatre, Leanard Huxley Building, Mills Road, ANU Campus. Andrew Freeman, who was Chairperson of OTSIG, and was Director, Community Affairs Board, ACS, Chaired this pre-conference workshop. Speakers included: *** ACTUAL AND POTENTIAL SUPPLIERS *** * STEVE WILLIAMS, WAS ACCOUNT MANAGER, IMAGING SYSTEMS, HERMES PRECISA AUSTRALIA (HPA), 'THE USE OF NEURAL NETWORK TECHNOLOGY AS AN AID FOR FORMS PROCESSING BY HPA'. Steve spoke on the use of neural network technology as an aid in processing education assessment forms in NSW and Victoria by HPA. He emphasised a number of aspects which need to be taken into account for such applications to be effective. These included, for example: * forms design being critical, possibly needing the assistance of a forms design expert; * cross-checking features needing to be incorporated in the forms; * pre-processing systems work needing to be undertaken in a planned fashion; and * comprehensive testing being critical to successful implementation. *** ACTUAL AND POTENTIAL USERS *** * ANDREW FREEMAN, WAS ASSISTANT DIRECTOR, MODE SECTION, DEPARTMENT OF IMMIGRATION AND MULTICURAL AFFAIRS (DIMA) ON 'THE POTENTIAL TO USE NEURAL NETWORK TECHNOLOGY AS AN AID IN PASSENGER CARD PROCESSING'. Passenger cards are filled in by all persons entering and leaving Australia. They are used for a range of purposes including: * statistical purposes; * as an aid in recording movements, where the traveller travels through a non-automated port (this relates only to around 1% of movements); * supporting evidence, to complement the computer record of movement. Much of the data on passenger cards is in either a mark sense format, or has character seperation. The would appear to be substantial potential to use neural network technology to assist with data capture from passenger cards. Some aspects which make this application challenging include: * the size of the passenger card, in relation to the amount of data recorded on it (the passenger card is A6 - 10.5 cm * 14.75 cm in area); and * the quality of the hand-print on the passenger cards is often highly variable. Examples of blank incoming and outgoing passenger cards were distributed at the workshop. * COMMENTS BY OTHER ATTENDEES AT THE WORKSHOP * There was general discussion about the Health Insurance Commission (HIC) application of neural network technology to assist with forms processing in that organisation. An attendee from the Attorney-General's Department (AGD) indicated he is particularly interested in how neural network technologies might be used as an aid in organising and distributing documentation (such as legal opinions) within an organisation.