Papers by Milica Gasic
Actor-Double-Critic: Incorporating Model-Based Critic for Task-Oriented Dialogue Systems (2020.findings-emnlp)
Copied to clipboard
| Challenge: | In order to improve the sample-efficiency of deep reinforcement learning, we implemented imagination augmented agent (I2A) in spoken dialogue systems (SDS). |
| Approach: | They propose to use an actor-double-critic to improve the stability and overall performance of imagination augmented agent (I2A) in spoken dialogue systems. |
| Outcome: | The proposed model-based agent (ADC) improves the stability and sample-efficiency of deep reinforcement learning (DRL) on a restaurant booking task. |
Knowing What You Know: Calibrating Dialogue Belief State Distributions via Ensembles (2020.findings-emnlp)
Copied to clipboard
Carel van Niekerk, Michael Heck, Christian Geishauser, Hsien-chin Lin, Nurul Lubis, Marco Moresi, Milica Gasic
| Challenge: | Current models for dialogue state tracking only achieve 55% accuracy . however, they lack in performance compared to belief trackers and do not produce well calibrated distributions. |
| Approach: | They propose to calibrate a model for dialogue belief trackers to measure dialogue state accuracy. |
| Outcome: | The proposed model outperforms existing models in terms of accuracy and accuracy. |
Speech-based Slot Filling using Large Language Models (2024.findings-acl)
Copied to clipboard
| Challenge: | Recent advances in large language models (LLMs) have shown an unprecedented ability across various language tasks. |
| Approach: | They propose to use prompts and LoRA fine-tuning to improve slot filling robustness . they propose a linearised knowledge injection scheme to integrate dynamic external knowledge into LLMs. |
| Outcome: | The proposed model improves slot filling with noisy ASR transcriptions with 6.7% and 17.6% absolute SLU-F1 improvements compared to a fully fine-tuned Flan-T5-XL model. |
ConvLab-3: A Flexible Dialogue System Toolkit Based on a Unified Data Format (2023.emnlp-demo)
Copied to clipboard
Qi Zhu, Christian Geishauser, Hsien-chin Lin, Carel van Niekerk, Baolin Peng, Zheng Zhang, Shutong Feng, Michael Heck, Nurul Lubis, Dazhen Wan, Xiaochen Zhu, Jianfeng Gao, Milica Gasic, Minlie Huang
| Challenge: | Existing tools for building TOD systems often lack a user-friendly interface . a toolkit with advanced, easily integrable modules is needed to bridge this gap . |
| Approach: | They propose a multifaceted dialogue system toolkit that integrates diverse datasets and models with a streamlined training process and in-depth evaluation tools. |
| Outcome: | The proposed toolkit combines RL and transfer learning to support the rapid development and evaluation of robust dialogue policies. |
Topology of Word Embeddings: Singularities Reflect Polysemy (2020.starsem-1)
Copied to clipboard
| Challenge: | a new study suggests that word vectors live on a submanifold within their ambient vector space . a manifold hypothesis suggests that vectors should live on pinched manifels . |
| Approach: | They propose a topological measure of polysemy that correlates well with the actual number of meanings of a word. |
| Outcome: | The proposed method produces competitive results for a word Sense Induction & Disambiguation task. |
Out-of-Task Training for Dialog State Tracking Models (2020.coling-main)
Copied to clipboard
Michael Heck, Christian Geishauser, Hsien-chin Lin, Nurul Lubis, Marco Moresi, Carel van Niekerk, Milica Gasic
| Challenge: | Dialog state tracking (DST) suffers from data sparsity. |
| Approach: | They utilize non-dialog data from unrelated NLP tasks to train dialog state trackers . they propose to use dialog state tracking to summarise the conversation history . |
| Outcome: | The proposed method exploits non-dialog data from unrelated NLP tasks to train dialog state trackers. |
EmoWOZ: A Large-Scale Corpus and Labelling Scheme for Emotion Recognition in Task-Oriented Dialogue Systems (2022.lrec-1)
Copied to clipboard
Shutong Feng, Nurul Lubis, Christian Geishauser, Hsien-chin Lin, Michael Heck, Carel van Niekerk, Milica Gasic
| Challenge: | Existing emotion-annotated task-oriented corpora are limited in size, label richness, and public availability, creating a bottleneck for downstream tasks. |
| Approach: | They propose a large-scale manually emotion-annotated corpus of task-oriented dialogues based on a multi-domain task-orientated dataset. |
| Outcome: | The proposed method is based on a task-oriented dialogue dataset with 11K dialogues and 83K emotion annotations of user utterances. |
ChatGPT for Zero-shot Dialogue State Tracking: A Solution or an Opportunity? (2023.acl-short)
Copied to clipboard
Michael Heck, Nurul Lubis, Benjamin Ruppik, Renato Vukovic, Shutong Feng, Christian Geishauser, Hsien-chin Lin, Carel van Niekerk, Milica Gasic
| Challenge: | Recent research on dialog state tracking (DST) focuses on methods that allow few- and zero-shot transfer to new domains or schemas. |
| Approach: | They propose to use schema descriptions to facilitate zero-shot transfer to new domains . they argue that general purpose language models lack the ability to replace specialized systems . |
| Outcome: | The proposed method achieves state-of-the-art in zero-shot DST with in-context learning capabilities. |
Uncertainty Measures in Neural Belief Tracking and the Effects on Dialogue Policy Performance (2021.emnlp-main)
Copied to clipboard
Carel van Niekerk, Andrey Malinin, Christian Geishauser, Michael Heck, Hsien-chin Lin, Nurul Lubis, Shutong Feng, Milica Gasic
| Challenge: | Neural dialogue belief trackers that take uncertainty into account are often overconfident in their decisions and therefore less robust. |
| Approach: | They propose to use different uncertainty measures in neural belief tracking to integrate uncertainty into the feature space of the policy and train policies through interaction with a user simulator. |
| Outcome: | The proposed approach improves both performance and robustness of the downstream dialogue policy. |
LAVA: Latent Action Spaces via Variational Auto-encoding for Dialogue Policy Optimization (2020.coling-main)
Copied to clipboard
Nurul Lubis, Christian Geishauser, Michael Heck, Hsien-chin Lin, Marco Moresi, Carel van Niekerk, Milica Gasic
| Challenge: | Reinforcement learning (RL) can be used to steer a conversation towards successful task completion. |
| Approach: | They propose to use latent latent variables to shape latent variable distributions . they use response auto-encoding as auxiliary task to capture generative factors . |
| Outcome: | The proposed approach yields a more action-characterized latent representations . the proposed approach achieves state-of-the-art success rates . |