arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1703.00607cs.CLstat.ML
收藏

动态词嵌入的语义发现

Dynamic Word Embeddings for Evolving Semantic Discovery

Rutgers University(罗格斯大学);Technicolor Research(特艺集团研究院);Amazon(亚马逊)

Zijun Yao, Yifan Sun, Weicong Ding, Nikhil Rao, Hui Xiong

摘要:

词汇进化是指随着时间的推移,词汇的含义和联想发生变化,这是人类语言进化的副产品。通过研究词汇的演变,我们可以推断出人类历史不同时期的社会趋势和语言结构。然而,传统的技术,如词表示学习不足以捕捉不断变化的语言结构和词汇。在本文中,我们开发了一个动态统计模型来学习时间感知的词向量表示。我们提出了一个模型,同时学习时间感知嵌入,并解决由此产生的“对齐问题”。该模型在抓取的NYTimes数据集上进行训练。此外,我们开发了多种直观的评价策略的时间词嵌入。我们的定性和定量测试表明,我们的方法不仅可靠地捕捉到这种随时间的演变,而且在语义准确性和对齐质量方面始终优于最先进的时间嵌入方法。

英文摘要:

Word evolution refers to the changing meanings and associations of words throughout time, as a byproduct of human language evolution. By studying word evolution, we can infer social trends and language constructs over different periods of human history. However, traditional techniques such as word representation learning do not adequately capture the evolving language structure and vocabulary. In this paper, we develop a dynamic statistical model to learn time-aware word vector representation. We propose a model that simultaneously learns time-aware embeddings and solves the resulting "alignment problem". This model is trained on a crawled NYTimes dataset. Additionally, we develop multiple intuitive evaluation strategies of temporal word embeddings. Our qualitative and quantitative tests indicate that our method not only reliably captures this evolution over time, but also consistently outperforms state-of-the-art temporal embedding approaches on both semantic accuracy and alignment quality.

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。

↑