Public Release: 

2,000 top experts in AI gather in Melbourne to discuss latest results and future directions

How will AI improve our lives? Can AI combat fake news? What challenges remain before machines match humans? Are killer robots a danger or a solution? Can AI play Angry Birds better than us?

University of New South Wales

A record 2,000 of the world's top experts in artificial intelligence (or AI) come together today in Melbourne for the 2017 International Joint Conference on Artificial Intelligence (IJCAI). Whilst many researchers believe 'thinking machines' that match humans are still 50 or more years away, there is considerable optimism for the challenge, as well as optimism about recent progress, reflected in the conference's growing size.

The conference, the leading venue internationally for AI research, attracted a record number of submissions and delegates, and there also are a record number of events: workshops, tutorials, scientific presentations and public talks.

"This year's conference theme is on the challenges of developing fully autonomous AI systems," said Carles Sierra, Professor at the Artificial Intelligence Research Institute (IIIA) of the Spanish Research Council in Barcelona, and Program Chair of the Conference.

"Large corporations and government agencies are developing autonomous systems that require our scientific position on sensitive legal and ethical questions. We will see in the following days how this theme pervades some of the invited talks, panels and specific sessions hopefully leading to meaningful answers by the end of the week.

"Our field is growing steadily thanks to new theoretical developments and a myriad of applications specially in the area of machine learning. This growth is particularly intense in China that is for the first time the leading country in terms of submissions and accepted papers. Another important highlight of this edition is the mobilisation of Australian researchers that contributed with eight percent of the papers and with more than 400 attendees."

Chinese participation in the conference is at an all-time high, representing one-quarter of all delegates. And one-third of scientific submissions to the conference came from China, more than from the United States and Europe put together. This reflects a growth spurt in AI research taking place in China, as well as the Chinese government's ambition to be the leading player in the field in the next decade.

Highlighted in the Monday's news conference will be two new and exciting applications of AI research. In a paper titled "A Convolutional Approach for Misinformation Identification" by Feng Yu, Qiang Liu, Shu Wu, Liang Wang and Tieniu Tan from the Chinese Academy of Sciences and China's National Laboratory of Pattern Recognition, the researchers describe how AI can be used to detect fake news.

Nowadays, many people access much of their news via social media, such as Twitter or Weibo. Unfortunately, social media provides fertile breeding ground for the spreading of misinformation, which can impact political elections and even pose threats to public security and social stability. To identify misinformation, a deep learning method is shown to be more effective than state-of-the-art methods. The method can detect most misinformation in the first 1-2 hours, which may be early enough to make an impact before it spreads too widely.

The second paper, "Tracking the Evolution of Customer Purchase Behavior Segmentation via a Fragmentation-Coagulation Process" by Ling Luo, Bin Li, Irena Koprinska, Shlomo Berkovsky and Fang Chen from the University of Sydney and CSIRO's Data61. This considers how a business can tailor commercial strategies in order to understand, attract and retain customers using emerging AI technologies for customer behaviour modelling. A case study was conducted with a major supermarket chain demonstrating that the methods can identify different customer groups, the triggers for customer purchase, and optimise the timing and focus of promotion campaigns.

To provide access to the conference to scientists from developing countries - and countries where obtaining appropriate visas are difficult to obtain - organisers will have eight tele-robots wandering the corridors and session rooms, allowing remote delegates to control the robots and watch the conference via a browser interface. The robots are provided by an Australian startup, Aubot founded by 2012 Young Australian of the Year, Marita Cheng. The eight robots are named after the capital cities of some of the countries most impacted by visa restrictions: Baghdad, Damascus, Khartoum, Sana'a, Tehran, Tripoli and Washington.

The conference will also host the annual Angry Birds AI Competition, AIBIRDS 2017. Its goal is to build computer programs which can play the popular game Angry Birds better than the best human players. On Thursday 24 August, the best AI programs will compete for the title of AIBIRDS Champion. The following day, the last of the conference, human competitors are invited to challenge the winning program to see if man or machine is mightier. In a recent survey among AI researchers, AIBIRDS has been named the AI challenge where AI is expected to beat humans next.

"Games like Angry Birds are surprisingly hard for computers to solve, as there are many possible actions and the outcome of each action is unknown in advance," said Jochen Renz, a professor at the Australian National University who is running the competition. "The techniques developed for this competition are important for building AI systems that can successfully interact with the physical world."

The conference is increasingly a venue to discuss the impact that technologies like AI will have on society, such as lethal autonomous weapons systems (also known as 'killer robots'), a topic that concerns AI and robotics researchers and will be one of the many social issues discussed at the conference.

Finally, every evening during the week, the conference will host a free panel discussion on the impact of AI on our lives, 'The AI Lounge', which is open to the public and produced jointly with the City of Melbourne's Melbourne Conversations. Attendees will be able to meet some of the world's leading AI experts and have questions answered about diverse topics, such as AI's impact on jobs, warfare or anything else. The conversation begins each night at 6pm in the Cargo Hall of the Melbourne Exhibition and Conference Centre. For a list of topics, see


Conference website:


The International Joint Conference on Artificial Intelligence (IJCAI) is the world's leading conference on artificial intelligence. It has been held every two years since 1969, and annually since 2015. It attracts around 2,000 of the best researchers working in AI from around the world. IJCAI 2017 is currently being held in Melbourne, Australia.

Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.