SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2009.11152 · EMNLP Findings · 2020

Hierarchical Pre-training for Sequence Labelling in Spoken Dialog

Matthieu Labeau, Matteo Manica, Pierre Colombo, Emile Chapuis, Chloe Clavel

arXiv · PDF · Open in the Atlas

Code that ran

We have not lifted any functions out of this paper's repositories yet, so there is nothing we have run. We have no repository linked to it either. That is a statement about our graph, not about the paper.

Abstract

Sequence labelling tasks like Dialog Act and Emotion/Sentiment identification are a key component of spoken dialog systems. In this work, we propose a new approach to learn generic representations adapted to spoken dialog, which we evaluate on a new benchmark we call Sequence labellIng evaLuatIon benChmark fOr spoken laN-guagE benchmark (SILICONE). SILICONE is model-agnostic and contains 10 di↵erent datasets of various sizes. We obtain our representations with a hierarchical encoder based on transformer architectures, for which we extend two well-known pretraining objectives. Pre-training is performed on OpenSubtitles: a large corpus of spoken dialog containing over 2.3 billion of tokens. We demonstrate how hierarchical encoders achieve competitive results with consistently fewer parameters compared to state-of-the-art models and we show their importance for both pre-training and finetuning.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2009.11152")
get_code_for_paper("2009.11152")
have("2009.11152")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2009.11152.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2009.11152.svg)](https://syntology.ai/paper/2009.11152)

Connect an agent — have() is free.