Computing machinery such as smartphones are ubiquitous, and so will be smart home appliances, self-driving cars and robots in the near future. Enabling these machines with natural language understanding abilities opens up potential opportunities for the broader society to benefit from, e.g., in accessing the world’s knowledge, or in controlling complex machines with little effort.

In this talk, we will focus on accessing knowledge stored in knowledge-bases and text documents in a colloquial manner. We will do so by processing language into programs that perform reasoning. The main questions we will explore are 1) which inductive biases on neural architectures are well suited for processing language to programs? 2) what is the role of linguistic structure? and 3) can we build models which produce answers without generating an explicit program?