Deep Learning and Reinforcement Learning research lab in Montreal
In January 2016, Maluuba opened a Deep Learning and Reinforcement Learning research lab in Montreal.
Our vision is a world where machines can reason and communicate with people to advance the collective intelligence of humanity. Maluuba Research focuses on two core problems in language-based AI: machine reading comprehension (MRC) and advanced conversational systems (ACS). Our goal here is to push fundamental breakthroughs in these domains.
Machine reading comprehension aims at teaching artificial agents to read and understand natural language. Almost all of human knowledge is stored in some form of text. From stone tablets to printed books to the internet, we write our ideas down. A machine capable of digesting and understanding this information would be incredibly useful. A machine that understands could answer users’ questions about arbitrary subjects in open domains like an oracle — and we wouldn’t have to engineer all useful information into some system in advance (an impossible task, anyhow). A machine that reads rapidly could further incorporate dynamic, ever-changing information like news and tweets on the fly.
But without effective natural-language interaction, the benefits of machines that read and comprehend will be largely inaccessible to most users. Most people aren’t programming experts who can query databases or parse replies. We want to empower users by democratizing access to, and control over, knowledgeable machines. That’s where advanced conversational systems come in. The role of conversation is to open an effective communication channel between knowledgeable machines and the typical user.
Maluuba Research is already making headway on these problems. We’ve published several academic papers on our MRC and ACS systems, and in the coming weeks we’ll be blogging here about the specific problems we’re tackling.
We’re also growing rapidly: We have 15 researchers, with new hires popping up monthly. Our machine-learning people come from all over the world — Europe, Asia, North America — and have published work at top conferences like ICML, NIPS, and ACL.
Of course, positioning ourselves in Montreal was no arbitrary choice. The city is a hotbed for Deep Learning research; most significantly, it’s the home of Yoshua Bengio and his MILA group at UdM. Our research lab has partnered with Professor Bengio with the aim of becoming the world’s most advanced research centre for language-based AI.
Stay tuned for more from Maluuba Research!