Modern neural sequence generation models are built to either generate tokens step-by-step from scratch or (iteratively) modify a sequence of tokens bounded by a fixed length. In this work, we develop Levenshtein Transformer, a new partially autoregressive model devised for more flexible and amenable sequence generation. Unlike previous approaches, the atomic operations of our model are insertion and deletion. The combination of them facilitates not only generation but also sequence refinement allowing dynamic length changes. We also propose a set of new training techniques dedicated at them, effectively exploiting one as the other's learning signal thanks to their complementary nature. Experiments applying the proposed model achieve comparable performance but much-improved efficiency on both generation (e.g. machine translation, text summarization) and refinement tasks (e.g. automatic post-editing). We further confirm the flexibility of our model by showing a Levenshtein Transformer trained by machine translation can straightforwardly be used for automatic post-editing.
Language model (LM) pre-training has resulted in impressive performance and sample efficiency on a variety of language understanding tasks. However, it remains unclear how to best use pre-trained LMs for generation tasks such as abstractive summarization, particularly to enhance sample efficiency. In these sequence-to-sequence settings, prior work has experimented with loading pre-trained weights into the encoder and/or decoder networks, but used non-pre-trained encoder-decoder attention weights. We instead use a pre-trained decoder-only network, where the same Transformer LM both encodes the source and generates the summary. This ensures that all parameters in the network, including those governing attention over source states, have been pre-trained before the fine-tuning step. Experiments on the CNN/Daily Mail dataset show that our pre-trained Transformer LM substantially improves over pre-trained Transformer encoder-decoder networks in limited-data settings. For instance, it achieves 13.1 ROUGE-2 using only 1% of the training data (~3000 examples), while pre-trained encoder-decoder models score 2.3 ROUGE-2.
This paper presents a new Unified pre-trained Language Model (UniLM) that can be fine-tuned for both natural language understanding and generation tasks. The model is pre-trained using three types of language modeling tasks: unidirectional, bidirectional, and sequence-to-sequence prediction. The unified modeling is achieved by employing a shared Transformer network and utilizing specific self-attention masks to control what context the prediction conditions on. UniLM compares favorably with BERT on the GLUE benchmark, and the SQuAD 2.0 and CoQA question answering tasks. Moreover, UniLM achieves new state-of-the-art results on five natural language generation datasets, including improving the CNN/DailyMail abstractive summarization ROUGE-L to 40.51 (2.04 absolute improvement), the Gigaword abstractive summarization ROUGE-L to 35.75 (0.86 absolute improvement), the CoQA generative question answering F1 score to 82.5 (37.1 absolute improvement), the SQuAD question generation BLEU-4 to 22.12 (3.75 absolute improvement), and the DSTC7 document-grounded dialog response generation NIST-4 to 2.67 (human performance is 2.65). The code and pre-trained models are available at https://github.com/microsoft/unilm.
As the amount of user-generated textual content grows rapidly, text summarization algorithms are increasingly being used to provide users a quick overview of the information content. Traditionally, summarization algorithms have been evaluated only based on how well they match human-written summaries (e.g. as measured by ROUGE scores). In this work, we propose to evaluate summarization algorithms from a completely new perspective that is important when the user-generated data to be summarized comes from different socially salient user groups, e.g. men or women, Caucasians or African-Americans, or different political groups (Republicans or Democrats). In such cases, we check whether the generated summaries fairly represent these different social groups. Specifically, considering that an extractive summarization algorithm selects a subset of the textual units (e.g. microblogs) in the original data for inclusion in the summary, we investigate whether this selection is fair or not. Our experiments over real-world microblog datasets show that existing summarization algorithms often represent the socially salient user-groups very differently compared to their distributions in the original data. More importantly, some groups are frequently under-represented in the generated summaries, and hence get far less exposure than what they would have obtained in the original data. To reduce such adverse impacts, we propose novel fairness-preserving summarization algorithms which produce high-quality summaries while ensuring fairness among various groups. To our knowledge, this is the first attempt to produce fair text summarization, and is likely to open up an interesting research direction.
Pre-training and fine-tuning, e.g., BERT, have achieved great success in language understanding by transferring knowledge from rich-resource pre-training task to the low/zero-resource downstream tasks. Inspired by the success of BERT, we propose MAsked Sequence to Sequence pre-training (MASS) for the encoder-decoder based language generation tasks. MASS adopts the encoder-decoder framework to reconstruct a sentence fragment given the remaining part of the sentence: its encoder takes a sentence with randomly masked fragment (several consecutive tokens) as input, and its decoder tries to predict this masked fragment. In this way, MASS can jointly train the encoder and decoder to develop the capability of representation extraction and language modeling. By further fine-tuning on a variety of zero/low-resource language generation tasks, including neural machine translation, text summarization and conversational response generation (3 tasks and totally 8 datasets), MASS achieves significant improvements over the baselines without pre-training or with other pre-training methods. Specially, we achieve the state-of-the-art accuracy (37.5 in terms of BLEU score) on the unsupervised English-French translation, even beating the early attention-based supervised model.
News headline generation is an essential problem of text summarization because it is constrained, well-defined, and is still hard to solve. Models with a limited vocabulary can not solve it well, as new named entities can appear regularly in the news and these entities often should be in the headline. News articles in morphologically rich languages such as Russian require model modifications due to a large number of possible word forms. This study aims to validate that models with a possibility of copying words from the original article performs better than models without such an option. The proposed model achieves a mean ROUGE score of 23 on the provided test dataset, which is 8 points greater than the result of a similar model without a copying mechanism. Moreover, the resulting model performs better than any known model on the new dataset of Russian news.
Neural encoder-decoder models have been successful in natural language generation tasks. However, real applications of abstractive summarization must consider additional constraint that a generated summary should not exceed a desired length. In this paper, we propose a simple but effective extension of a sinusoidal positional encoding (Vaswani et al., 2017) to enable neural encoder-decoder model to preserves the length constraint. Unlike in previous studies where that learn embeddings representing each length, the proposed method can generate a text of any length even if the target length is not present in training data. The experimental results show that the proposed method can not only control the generation length but also improve the ROUGE scores.
We present a new neural model for text summarization that first extracts sentences from a document and then compresses them. The proposed model offers a balance that sidesteps the difficulties in abstractive methods while generating more concise summaries than extractive methods. In addition, our model dynamically determines the length of the output summary based on the gold summaries it observes during training and does not require length constraints typical to extractive summarization. The model achieves state-of-the-art results on the CNN/DailyMail and Newsroom datasets, improving over current extractive and abstractive methods. Human evaluations demonstrate that our model generates concise and informative summaries. We also make available a new dataset of oracle compressive summaries derived automatically from the CNN/DailyMail reference summaries.
Recent neural network approaches to summarization are largely either selection-based extraction or generation-based abstraction. In this work, we present a neural model for single-document summarization based on joint extraction and syntactic compression. Our model chooses sentences from the document, identifies possible compressions based on constituency parses, and scores those compressions with a neural model to produce the final summary. For learning, we construct oracle extractive-compressive summaries, then learn both of our components jointly with this supervision. Experimental results on the CNN/Daily Mail and New York Times datasets show that our model achieves strong performance (comparable to state-of-the-art systems) as evaluated by ROUGE. Moreover, our approach outperforms an off-the-shelf compression module, and human and manual evaluation shows that our model's output generally remains grammatical.
In the past few years, neural abstractive text summarization with sequence-to-sequence (seq2seq) models have gained a lot of popularity. Many interesting techniques have been proposed to improve seq2seq models, making them capable of handling different challenges, such as saliency, fluency and human readability, and generate high-quality summaries. Generally speaking, most of these techniques differ in one of these three categories: network structure, parameter inference, and decoding/generation. There are also other concerns, such as efficiency and parallelism for training a model. In this paper, we provide a comprehensive literature survey on different seq2seq models for abstractive text summarization from the viewpoint of network structures, training strategies, and summary generation algorithms. Several models were first proposed for language modeling and generation tasks, such as machine translation, and later applied to abstractive text summarization. Hence, we also provide a brief review of these models. As part of this survey, we also develop an open source library, namely, Neural Abstractive Text Summarizer (NATS) toolkit, for the abstractive text summarization. An extensive set of experiments have been conducted on the widely used CNN/Daily Mail dataset to examine the effectiveness of several different neural network components. Finally, we benchmark two models implemented in NATS on the two recently released datasets, namely, Newsroom and Bytecup.
Deep neural networks are data hungry models and thus face difficulties when attempting to train on small text datasets. Transfer learning is a potential solution but their effectiveness in the text domain is not as explored as in areas such as image analysis. In this paper, we study the problem of transfer learning for text summarization and discuss why existing state-of-the-art models fail to generalize well on other (unseen) datasets. We propose a reinforcement learning framework based on a self-critic policy gradient approach which achieves good generalization and state-of-the-art results on a variety of datasets. Through an extensive set of experiments, we also show the ability of our proposed framework to fine-tune the text summarization model using only a few training samples. To the best of our knowledge, this is the first work that studies transfer learning in text summarization and provides a generic solution that works well on unseen data.
Automatic text summarization is generally considered as a challenging task in the NLP community. One of the challenges is the publicly available and large dataset that is relatively rare and difficult to construct. The problem is even worse for low-resource languages such as Indonesian. In this paper, we present IndoSum, a new benchmark dataset for Indonesian text summarization. The dataset consists of news articles and manually constructed summaries. Notably, the dataset is almost 200x larger than the previous Indonesian summarization dataset of the same domain. We evaluated various extractive summarization approaches and obtained encouraging results which demonstrate the usefulness of the dataset and provide baselines for future research. The code and the dataset are available online under permissive licenses.
In this paper, we introduce Iterative Text Summarization (ITS), an iteration-based model for supervised extractive text summarization, inspired by the observation that it is often necessary for a human to read an article multiple times in order to fully understand and summarize its contents. Current summarization approaches read through a document only once to generate a document representation, resulting in a sub-optimal representation. To address this issue we introduce a model which iteratively polishes the document representation on many passes through the document. As part of our model, we also introduce a selective reading mechanism that decides more accurately the extent to which each sentence in the model should be updated. Experimental results on the CNN/DailyMail and DUC2002 datasets demonstrate that our model significantly outperforms state-of-the-art extractive systems when evaluated by machines and by humans.
In sentence compression, the task of shortening sentences while retaining the original meaning, models tend to be trained on large corpora containing pairs of verbose and compressed sentences. To remove the need for paired corpora, we emulate a summarization task and add noise to extend sentences and train a denoising auto-encoder to recover the original, constructing an end-to-end training regime without the need for any examples of compressed sentences. We conduct a human evaluation of our model on a standard text summarization dataset and show that it performs comparably to a supervised baseline based on grammatical correctness and retention of meaning. Despite being exposed to no target data, our unsupervised models learn to generate imperfect but reasonably readable sentence summaries. Although we underperform supervised models based on ROUGE scores, our models are competitive with a supervised baseline based on human evaluation for grammatical correctness and retention of meaning.
We introduce extreme summarization, a new single-document summarization task which does not favor extractive strategies and calls for an abstractive modeling approach. The idea is to create a short, one-sentence news summary answering the question "What is the article about?". We collect a real-world, large-scale dataset for this task by harvesting online articles from the British Broadcasting Corporation (BBC). We propose a novel abstractive model which is conditioned on the article's topics and based entirely on convolutional neural networks. We demonstrate experimentally that this architecture captures long-range dependencies in a document and recognizes pertinent content, outperforming an oracle extractive system and state-of-the-art abstractive approaches when evaluated automatically and by humans.
ROUGE is a widely adopted, automatic evaluation measure for text summarization. While it has been shown to correlate well with human judgements, it is biased towards surface lexical similarities. This makes it unsuitable for the evaluation of abstractive summarization, or summaries with substantial paraphrasing. We study the effectiveness of word embeddings to overcome this disadvantage of ROUGE. Specifically, instead of measuring lexical overlaps, word embeddings are used to compute the semantic similarity of the words used in summaries instead. Our experimental results show that our proposal is able to achieve better correlations with human judgements when measured with the Spearman and Kendall rank coefficients.
Neural encoder-decoder models have shown great success in many sequence generation tasks. However, previous work has not investigated situations in which we would like to control the length of encoder-decoder outputs. This capability is crucial for applications such as text summarization, in which we have to generate concise summaries with a desired length. In this paper, we propose methods for controlling the output sequence length for neural encoder-decoder models: two decoding-based methods and two learning-based methods. Results show that our learning-based methods have the capability to control length without degrading summary quality in a summarization task.
Neural sequence-to-sequence models have provided a viable new approach for abstractive text summarization (meaning they are not restricted to simply selecting and rearranging passages from the original text). However, these models have two shortcomings: they are liable to reproduce factual details inaccurately, and they tend to repeat themselves. In this work we propose a novel architecture that augments the standard sequence-to-sequence attentional model in two orthogonal ways. First, we use a hybrid pointer-generator network that can copy words from the source text via pointing, which aids accurate reproduction of information, while retaining the ability to produce novel words through the generator. Second, we use coverage to keep track of what has been summarized, which discourages repetition. We apply our model to the CNN / Daily Mail summarization task, outperforming the current abstractive state-of-the-art by at least 2 ROUGE points.
Recurrent neural network models with an attention mechanism have proven to be extremely effective on a wide variety of sequence-to-sequence problems. However, the fact that soft attention mechanisms perform a pass over the entire input sequence when producing each element in the output sequence precludes their use in online settings and results in a quadratic time complexity. Based on the insight that the alignment between input and output sequence elements is monotonic in many problems of interest, we propose an end-to-end differentiable method for learning monotonic alignments which, at test time, enables computing attention online and in linear time. We validate our approach on sentence summarization, machine translation, and online speech recognition problems and achieve results competitive with existing sequence-to-sequence models.
This paper proposes a text summarization approach for factual reports using a deep learning model. This approach consists of three phases: feature extraction, feature enhancement, and summary generation, which work together to assimilate core information and generate a coherent, understandable summary. We are exploring various features to improve the set of sentences selected for the summary, and are using a Restricted Boltzmann Machine to enhance and abstract those features to improve resultant accuracy without losing any important information. The sentences are scored based on those enhanced features and an extractive summary is constructed. Experimentation carried out on several articles demonstrates the effectiveness of the proposed approach. Source code available at: https://github.com/vagisha-nidhi/TextSummarizer