arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2006.13299cs.CLcs.LG
收藏

词嵌入的监督理解

Supervised Understanding of Word Embeddings

Siemens Medical Solutions USA(美国西门子医疗解决方案公司)

Halid Ziya Yerebakan, Parmeet Bhatia, Yoshihisa Shinagawa

摘要:

预训练的词嵌入被广泛用于自然语言处理中的迁移学习。嵌入是单词的连续和分布式表示,在紧凑的欧几里得空间中保持它们的相似性。然而,这些空间的尺寸并没有提供任何明确的解释。在这项研究中,我们已经获得了监督投影的形式线性关键字级分类器的词嵌入。我们已经表明,该方法创建原始嵌入尺寸的可解释的投影。经训练的分类器节点的激活对应于词汇表中的单词的子集。因此,它们的行为类似于字典特征,同时具有连续值输出的优点。此外,这样的字典可以通过将得分最高的单词上的专家标签添加到关键字的初始集合来迭代地增长多轮。此外,相同的分类器可以应用于其他语言中的对齐的词嵌入以获得对应的词典。在我们的实验中,我们已经证明,使用这些分类器权重初始化高阶网络可以为下游NLP任务提供更准确的模型。我们进一步证明了监督维度在揭示感兴趣的关键词的多义性质方面的有用性,通过在不同的子空间中使用学习的分类器来投影它的嵌入。

英文摘要:

Pre-trained word embeddings are widely used for transfer learning in natural language processing. The embeddings are continuous and distributed representations of the words that preserve their similarities in compact Euclidean spaces. However, the dimensions of these spaces do not provide any clear interpretation. In this study, we have obtained supervised projections in the form of the linear keyword-level classifiers on word embeddings. We have shown that the method creates interpretable projections of original embedding dimensions. Activations of the trained classifier nodes correspond to a subset of the words in the vocabulary. Thus, they behave similarly to the dictionary features while having the merit of continuous value output. Additionally, such dictionaries can be grown iteratively with multiple rounds by adding expert labels on top-scoring words to an initial collection of the keywords. Also, the same classifiers can be applied to aligned word embeddings in other languages to obtain corresponding dictionaries. In our experiments, we have shown that initializing higher-order networks with these classifier weights gives more accurate models for downstream NLP tasks. We further demonstrate the usefulness of supervised dimensions in revealing the polysemous nature of a keyword of interest by projecting it's embedding using learned classifiers in different sub-spaces.

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。

↑