portal for the AISB 2019 Symposium on Language Learning for Artificial Agents
This symposium has come about through the culmination of one project, and the inception of another.
Atlantis is a CHIST-ERA funded project focusing on applying theoretical explorations of the earliest stages of language learning to the development of artificial linguistic agents. This research has entailed the generation of new datasets, the modelling of the way environmental affordances become the semiotic substrate for the formation of composable semantic representations, and hands-on experiments involving robots. In addition to the organisations represented on the organising committee of this symposium (Lattice, OFAI, and the VUB Artificial Intelligence Lab), the project includes researchers from the Institut de Biologia Evolutiva and Sony Labs.
Ralli is a WWTF funded project exploring the relationship between task-oriented language and social robotics. Motivated by insight into the fundamental role of language in coordinating action in an environment, the project seeks to model the way that artificial agents might learn from multimodal interaction with human tutors in task-oriented situations. The project consortium consists of OFAI, ACIN (TU Vienna) and HRI Lab (Tufts University).