arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2010.04529cs.CL
收藏

我们在文本摘要方面取得了哪些成就?

What Have We Achieved on Text Summarization?

Westlake University(西湖大学);Westlake Institute for Advanced Study(西湖高等研究院);Zhejiang University(浙江大学);Tencent SPPD(腾讯可持续社会价值事业部)

Dandan Huang, Leyang Cui, Sen Yang, Guangsheng Bao, Kun Wang, Jun Xie, Yue Zhang

摘要:

多年来,通过研究各种方法并提高报告的ROUGE分数,深度学习已经显着改善了文本摘要。然而,自动摘要器和人类专业人员制作的摘要之间仍然存在差距。为了更好地了解摘要系统在细粒度语法和语义层面上的优势和局限性,我们参考了多维质量度量(MQM),并手动量化了10个代表性摘要模型的8个主要错误来源。首先,我们发现:1)在相似的设置下,由于忠实性和事实一致性的优势,提取摘要通常优于抽象摘要; 2)复制,覆盖和混合提取/抽象方法等里程碑技术确实带来了具体的改进,但也表现出局限性; 3)预训练技术,特别是序列到序列的预训练,对于改进文本摘要非常有效,BART给出了最好的结果。

英文摘要:

Deep learning has led to significant improvement in text summarization with various methods investigated and improved ROUGE scores reported over the years. However, gaps still exist between summaries produced by automatic summarizers and human professionals. Aiming to gain more understanding of summarization systems with respect to their strengths and limits on a fine-grained syntactic and semantic level, we consult the Multidimensional Quality Metric(MQM) and quantify 8 major sources of errors on 10 representative summarization models manually. Primarily, we find that 1) under similar settings, extractive summarizers are in general better than their abstractive counterparts thanks to strength in faithfulness and factual-consistency; 2) milestone techniques such as copy, coverage and hybrid extractive/abstractive methods do bring specific improvements but also demonstrate limitations; 3) pre-training techniques, and in particular sequence-to-sequence pre-training, are highly effective for improving text summarization, with BART giving the best results.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑