Few-Shot Generalization Across Dialogue Tasks

Vladimir Vlasov, Akela Drissner-Schmid, Alan Nichol
11/28/2018
cs.CLcs.AI

Official code: RasaHQ/conversational-ai-workshop-18 (68 stars, last commit 2020-08-30).

Abstract

Machine-learning based dialogue managers are able to learn complex behaviors in order to complete a task, but it is not straightforward to extend their capabilities to new domains. We investigate different policies' ability to handle uncooperative user behavior, and how well expertise in completing one task (such as restaurant reservations) can be reapplied when learning a new one (e.g. booking a hotel). We introduce the Recurrent Embedding Dialogue Policy (REDP), which embeds system actions and dialogue states in the same vector space. REDP contains a memory component and attention mechanism based on a modified Neural Turing Machine, and significantly outperforms a baseline LSTM classifier on this task. We also show that both our architecture and baseline solve the bAbI dialogue task, achieving 100% test accuracy.

View on arXivView PDF

Code Implementations(1)

RasaHQ/conversational-ai-workshop-18Official100%

Example showing generalisation

6829Python, MakefileMar 28, 20186 years ago

Similar Papers

OmniDialog: An Omnipotent Pre-training Model for Task-Oriented Dialogue System

83% match1 repo1622023

End-to-End Task-Completion Neural Dialogue Systems

83% match1 repo8032017

Latent-IM: Latent Interaction Management for Speech LLMs

83% match2026

Deep Dyna-Q: Integrating Planning for Task-Completion Dialogue Policy Learning

82% match1 repo1522018

CrossWOZ: A Large-Scale Chinese Cross-Domain Task-Oriented Dialogue Dataset

82% match1 repo7252020

Act-Aware Slot-Value Predicting in Multi-Domain Dialogue State Tracking

82% match1 repo42022

Cite this paper

@article{vlasov2018fewshot,
  title  = {Few-Shot Generalization Across Dialogue Tasks},
  author = {Vladimir Vlasov and Akela Drissner-Schmid and Alan Nichol},
  year   = {2018},
  eprint = {1811.11707},
  archivePrefix = {arXiv},
  url    = {https://arxiv.org/abs/1811.11707}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/1811.11707.svg)](https://codeofpaper.com/paper/1811.11707)

Discussion