ACL is the premier conference in the field of computational linguistics, covering a broad spectrum of diverse research areas that are concerned with computational approaches to natural language.
The 57th Annual Meeting of the Association for Computational Linguistics (ACL) is being held in Florence (Italy) at the ‘Fortezza da Basso‘. Today is the last day of this 6-day event and here are some papers that took the top honours:
Zero-Shot Word Sense Disambiguation Using Sense Definition Embeddings via IISc Bangalore & CMU
Word Sense Disambiguation (WSD) is a longstanding but open problem in Natural Language Processing (NLP).
Current supervised WSD methods treat senses as discrete labels and also resort to predicting the Most-Frequent-Sense (MFS) for words unseen during training.
The researchers from IISc Bangalore in collaboration with Carnegie Mellon University propose Extended WSD Incorporating Sense Embeddings (EWISE), a supervised model to perform WSD by predicting over a continuous sense embedding space as opposed to a discrete label space.
Check the full paper here.
Bridging the Gap Between Training and Inference for Neural Machine Translation via Chinese Academy Of Sciences et al.,
Neural Machine Translation (NMT) generates target words sequentially in the way of predicting the next word conditioned on the context words. At training time, it predicts with the ground truth words as context while at inference it has to generate the entire sequence from scratch. This discrepancy of the fed context leads to error accumulation along the way.
In this paper, the authors address these issues by sampling context words not only from the ground truth sequence but also from the predicted sequence by the model during training, where the predicted sequence is selected with a sentence-level optimum. Experiment results on Chinese->English and WMT’14 English->German translation tasks demonstrate that this approach can achieve significant improvements on multiple datasets.
Check the paper here.
Emotion-Cause Pair Extraction: A New Task to Emotion Analysis In Texts via Nanjing University of Science and Technology, China
Emotion cause extraction (ECE), the task aimed at extracting the potential causes behind certain emotions in text has gained much attention in recent years due to its wide applications.
In this work, the authors propose a new task: emotion-cause pair extraction (ECPE), which aims to extract the potential pairs of emotions and corresponding causes in a document. They propose a 2-step approach to address this new ECPE task, which first performs individual emotion extraction and cause extraction via multi-task learning, and then conduct emotion-cause pairing and filtering. The experimental results on a benchmark emotion cause corpus prove the feasibility of the ECPE task as well as the effectiveness of this approach.
Check this paper here
A Simple Theoretical Model of Importance for Summarization From via Swiss Federal Institute of Technology Lausanne, Switzerland
The author proposes a simple but rigorous definition of several concepts that were previously used only intuitively in summarization: Redundancy, Relevance, and Informativeness.Importance arises as a single quantity naturally unifying these concepts. Additionally, the author Maxim Peyrard provides intuition to interpret the proposed quantities and experiments to demonstrate the potential of the framework to inform and guide subsequent works.
Check the paper here.
Transferable Multi-Domain State Generator for Task-Oriented Dialogue Systems via The Hong Kong University of Science and Technology & Salesforce Research
Existing approaches generally fall short in tracking unknown slot values during inference and often have difficulties in adapting to new domains. In this paper, the authors propose a Transferable Dialogue State Generator (TRADE) that generates dialogue states from utterances using a copy mechanism, facilitating knowledge transfer when predicting (domain, slot, value) triplets not encountered during training.
TRADE achieves 60.58% joint goal accuracy in one of the zero-shot domains, and is able to adapt to few-shot cases without forgetting already trained domains.
Check the paper here.
We Need To Talk About Standard Splits via City University Of NewYork & Oregon Health & Science University
Few researchers apply statistical tests to determine whether differences in performance are likely to arise by chance, and few examine the stability of system ranking across multiple training-testing splits. In this paper, the authors conduct replication and reproduction experiments with nine part-of-speech taggers published between 2000 and 2018, each of which reports state-of-the-art performance on a widely-used “standard split”.
They fail to reliably reproduce some rankings using randomly generated splits. In this paper, the authors suggest that randomly generated splits should be used in system comparison.
Check the paper here.