Google for Social Good - Using Artificial Intelligence (AI)

Overview

This AI project was funded as one of 20 projects from among 2600 applicants from around the world for the use of AI for Social Impact. The project utilises the data coded as part of NASS to develop machine learning techniques that will enable automated coding to augment the work of human coders.

Over time, the volume and complexity of the electronic patient care records derived from ambulance attendances have increased. AI technology has the capacity to filter cases that do not meet project inclusion criteria, and automate coding of less complex cases so that coders can focus on more complex coding tasks.

These new methodologies are intended to improve not only the efficiency and accuracy of the coded database, but awareness and reporting of trends and characteristics relating to AOD use, self-harm, and violence in Australia.

The project has three specific aims:

  1. Develop and test an AI algorithm to filter non-relevant cases

  2. Implement the AI algorithm to code data

  3. Develop a new coding interface that improves coding experience and efficiency

Project team

Professor Dan Lubman, Dr Debbie Scott, Professor Wray Buntine, and Mr Sam Campbell.

Our funder

Google.org