128 papers with code • 4 benchmarks • 23 datasets
Automatic Document Summarization is the task of rewriting a document into its shorter form while still retaining its important content. The most popular two paradigms are extractive approaches and abstractive approaches. Extractive approaches generate summaries by extracting parts of the original document (usually sentences), while abstractive methods may generate new words or phrases which are not in the original document.
Natural language processing tasks, such as question answering, machine translation, reading comprehension, and summarization, are typically approached with supervised learning on taskspecific datasets.
Ranked #1 on Language Modelling on enwik8 (using extra training data)
This paper presents a new Unified pre-trained Language Model (UniLM) that can be fine-tuned for both natural language understanding and generation tasks.
Ranked #2 on Generative Question Answering on CoQA (using extra training data)
Comparing with traditional methods, our method has two main advantages: (1) the relations between sentences are captured by modeling both the graph structure of the whole document set and the candidate sub-graphs; (2) directly outputs an integrate summary in the form of sub-graph which is more informative and coherent.
Graphs that capture relations between textual units have great benefits for detecting salient information from multiple documents and generating overall coherent summaries.
For abstractive summarization, we propose a new fine-tuning schedule which adopts different optimizers for the encoder and the decoder as a means of alleviating the mismatch between the two (the former is pretrained while the latter is not).
Ranked #4 on Text Summarization on X-Sum
In this paper we conceptualize extractive summarization as a sentence ranking task and propose a novel training algorithm which globally optimizes the ROUGE evaluation metric through a reinforcement learning objective.
Ranked #12 on Extractive Text Summarization on CNN / Daily Mail
We introduce 'extreme summarization', a new single-document summarization task which aims at creating a short, one-sentence news summary answering the question ``What is the article about?''.