Towards an Automated Language Acquisition System for Grounded Agency

The Automated Language Acquisition System focuses on the fundamental question of grounding; that is, how an agent acquires and represents the meaning of concepts. We take the view that prior to acquiring natural language, an agent has experienced the world in a largely private manner. The agent has visual experiences from which it has learned that objects and object categories exist, they persist over time and possess attributes. The agent also understands that events take place in a physical world and that agents exist and have purpose. We show that Emergent Languages, which can be constructed in an unsupervised manner, can be thought of as a private language. A natural language such as English can then be viewed as a repository of concepts. By applying unsupervised clustering to images described via an Emergent Language, we show that an agent can then analyze such clusters and map them to their associated natural language concepts with the aid of a Natural Language expert. Another form of experience is exposure to spoken speech, the idea being that syntax and statistical frequencies can be observed. This allows for a form of inductive learning where an agent can expand its conceptual knowledge. Having discovered the existence of novel concepts and forming mappings to associated Emergent Language description through machine translation, relationships between concepts are then constructed. The hypotheses being that the meaning of a concept emerges from its connections with other concepts. Relationships that are investigated include: physical and causal, metaphorical and pragmatic.