A Multi-Dimensional, Cross-Domain and Hierarchy-Aware Neural Architecture for ISO-Standard Dialogue Act Tagging

Stefano Mezza, Wayne Wobcke, Alan Blair

Abstract:

Dialogue Act tagging with the ISO 24617-2
standard is a difficult task that involves multi-
label text classification across a diverse set of la-
bels covering semantic, syntactic and pragmatic
aspects of dialogue. The lack of an adequately
sized training set annotated with this taxonomy
is a major problem when using the standard
in practice. In this work, we propose a neural
architecture to increase classification accuracy,
especially on low-frequency fine-grained tags,
on a subset of the ISO 24617-2 taxonomy. Our
model takes advantage of the hierarchical struc-
ture of the ISO taxonomy and utilises syntactic
information in the form of Part-Of-Speech and
dependency tags, in addition to contextual in-
formation from previous turns. We train our
architecture on an aggregated corpus of conver-
sations from different domains, which provides
a variety of dialogue interactions and linguistic
registers. Our approach achieves state-of-the-
art tagging results on the DialogBank bench-
mark data set, providing empirical evidence
that this architecture can successfully gener-
alise to different domains.