IJCAI 2019 in Macau, China
In August 2019, Andrew attended the 28th International Joint Conference on Artificial Intelligence (IJCAI-2019). He shares his learning and experience in this blog post.
IJCAI 2019 was hosted in Macau, China, gathering more than 3000 AI scholars and industry leaders. This year, IJCAI received a huge amount of submissions, i.e., 4752, which resulted in 17.8% acceptance rate (850 papers). This is a record for IJCAI, showing a rapidly increasing interest in AI.
IJCAI 2019 started with many activities: tutorials, workshops, demos, followed by the main conference sessions. On the first day of the conference, participants were treated with an amazing Kylin dance performance from the local artists. This signaled a very fruitful, educational, and eye-opening experience for me.
In IJCAI 2019, Andrew presented his paper on “Learning Multiple Maps from Conditional Ordinal Triplets“, introducing the problem of modelling multiple similarity perspectives from the similarity triplets. The paper also proposes a geometric framework to effectively model multiple similarity perspectives from the ordinal similarity triplets. This aims at learning multiple low-dimensional representations from such similarity triplets, each reflecting a specific similarity perspective.
Preferred.AI also has another paper in IJCAI 2019, titled “Correlation-Sensitive Next-Basket Recommendation, discussing the next-basket recommendation problem. Motivated by the observation that items are commonly adopted concurrently, e.g., a basket of grocery items or a sitting of media consumption, this paper deals with a sequence of baskets as input, and seeks to recommend the next basket.
IJCAI 2019 was also an opportunity for conference attendants to explore Macau, a special administrative zone of China. Macau is famous for many huge and luxury hotels with a countless number of casinos. Visitors can easily find themselves in a casino just by walking from hotels to hotels.
IJCAI is always a great venue to meet and learn from world-class AI researchers and practitioners. Preferred.AI will definitely be back again!