This paper proposes a novel architecture, Cross Attention Augmented Transducer (CAAT), for simultaneous translation. The framework aims to jointly optimize the policy and translation models. To effectively consider all possible READ-WRITE simultaneou
s translation action paths, we adapt the online automatic speech recognition (ASR) model, RNN-T, but remove the strong monotonic constraint, which is critical for the translation task to consider reordering. To make CAAT work, we introduce a novel latency loss whose expectation can be optimized by a forward-backward algorithm. We implement CAAT with Transformer while the general CAAT architecture can also be implemented with other attention-based encoder-decoder frameworks. Experiments on both speech-to-text (S2T) and text-to-text (T2T) simultaneous translation tasks show that CAAT achieves significantly better latency-quality trade-offs compared to the state-of-the-art simultaneous translation approaches.
Keyword or keyphrase extraction is to identify words or phrases presenting the main topics of a document. This paper proposes the AttentionRank, a hybrid attention model, to identify keyphrases from a document in an unsupervised manner. AttentionRank
calculates self-attention and cross-attention using a pre-trained language model. The self-attention is designed to determine the importance of a candidate within the context of a sentence. The cross-attention is calculated to identify the semantic relevance between a candidate and sentences within a document. We evaluate the AttentionRank on three publicly available datasets against seven baselines. The results show that the AttentionRank is an effective and robust unsupervised keyphrase extraction model on both long and short documents. Source code is available on Github.
The Transformer translation model is based on the multi-head attention mechanism, which can be parallelized easily. The multi-head attention network performs the scaled dot-product attention function in parallel, empowering the model by jointly atten
ding to information from different representation subspaces at different positions. In this paper, we present an approach to learning a hard retrieval attention where an attention head only attends to one token in the sentence rather than all tokens. The matrix multiplication between attention probabilities and the value sequence in the standard scaled dot-product attention can thus be replaced by a simple and efficient retrieval operation. We show that our hard retrieval attention mechanism is 1.43 times faster in decoding, while preserving translation quality on a wide range of machine translation tasks when used in the decoder self- and cross-attention networks.
Recently, it has been argued that encoder-decoder models can be made more interpretable by replacing the softmax function in the attention with its sparse variants. In this work, we introduce a novel, simple method for achieving sparsity in attention
: we replace the softmax activation with a ReLU, and show that sparsity naturally emerges from such a formulation. Training stability is achieved with layer normalization with either a specialized initialization or an additional gating function. Our model, which we call Rectified Linear Attention (ReLA), is easy to implement and more efficient than previously proposed sparse attention mechanisms. We apply ReLA to the Transformer and conduct experiments on five machine translation tasks. ReLA achieves translation performance comparable to several strong baselines, with training and decoding speed similar to that of the vanilla attention. Our analysis shows that ReLA delivers high sparsity rate and head diversity, and the induced cross attention achieves better accuracy with respect to source-target word alignment than recent sparsified softmax-based models. Intriguingly, ReLA heads also learn to attend to nothing (i.e. switch off') for some queries, which is not possible with sparsified softmax alternatives.
In this work, we conduct a comprehensive investigation on one of the centerpieces of modern machine translation systems: the encoder-decoder attention mechanism. Motivated by the concept of first-order alignments, we extend the (cross-)attention mech
anism by a recurrent connection, allowing direct access to previous attention/alignment decisions. We propose several ways to include such a recurrency into the attention mechanism. Verifying their performance across different translation tasks we conclude that these extensions and dependencies are not beneficial for the translation performance of the Transformer architecture.
We use Hypergraph Attention Networks (HyperGAT) to recognize multiple labels of Chinese humor texts. We firstly represent a joke as a hypergraph. The sequential hyperedge and semantic hyperedge structures are used to construct hyperedges. Then, atten
tion mechanisms are adopted to aggregate context information embedded in nodes and hyperedges. Finally, we use trained HyperGAT to complete the multi-label classification task. Experimental results on the Chinese humor multi-label dataset showed that HyperGAT model outperforms previous sequence-based (CNN, BiLSTM, FastText) and graph-based (Graph-CNN, TextGCN, Text Level GNN) deep learning models.
We propose a novel framework for predicting the factuality of reporting of news media outlets by studying the user attention cycles in their YouTube channels. In particular, we design a rich set of features derived from the temporal evolution of the
number of views, likes, dislikes, and comments for a video, which we then aggregate to the channel level. We develop and release a dataset for the task, containing observations of user attention on YouTube channels for 489 news media. Our experiments demonstrate both complementarity and sizable improvements over state-of-the-art textual representations.
In this paper, we describe our system used for SemEval 2021 Task 5: Toxic Spans Detection. Our proposed system approaches the problem as a token classification task. We trained our model to find toxic words and concatenate their spans to predict the
toxic spans within a sentence. We fine-tuned Pre-trained Language Models (PLMs) for identifying the toxic words. For fine-tuning, we stacked the classification layer on top of the PLM features of each word to classify if it is toxic or not. PLMs are pre-trained using different objectives and their performance may differ on downstream tasks. We, therefore, compare the performance of BERT, ELECTRA, RoBERTa, XLM-RoBERTa, T5, XLNet, and MPNet for identifying toxic spans within a sentence. Our best performing system used RoBERTa. It performed well, achieving an F1 score of 0.6841 and secured a rank of 16 on the official leaderboard.
Toxic Spans Detection(TSD) task is defined as highlighting spans that make a text toxic. Many works have been done to classify a given comment or document as toxic or non-toxic. However, none of those proposed models work at the token level. In this
paper, we propose a self-attention-based bidirectional gated recurrent unit(BiGRU) with a multi-embedding representation of the tokens. Our proposed model enriches the representation by a combination of GPT-2, GloVe, and RoBERTa embeddings, which led to promising results. Experimental results show that our proposed approach is very effective in detecting span tokens.
This paper presents our submission to SemEval-2021 Task 5: Toxic Spans Detection. The purpose of this task is to detect the spans that make a text toxic, which is a complex labour for several reasons. Firstly, because of the intrinsic subjectivity of
toxicity, and secondly, due to toxicity not always coming from single words like insults or offends, but sometimes from whole expressions formed by words that may not be toxic individually. Following this idea of focusing on both single words and multi-word expressions, we study the impact of using a multi-depth DistilBERT model, which uses embeddings from different layers to estimate the final per-token toxicity. Our quantitative results show that using information from multiple depths boosts the performance of the model. Finally, we also analyze our best model qualitatively.
