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 1506.02004cs.CL
收藏

稀疏过完备词向量表示

Sparse Overcomplete Word Vector Representations

Carnegie Mellon University(卡内基梅隆大学)

Manaal Faruqui, Yulia Tsvetkov, Dani Yogatama, Chris Dyer, Noah Smith

摘要:

当前单词的分布式表示与词汇语义理论几乎没有相似之处。前者是密集和不可解释的,后者主要基于熟悉的离散类(例如,超意义)和关系(例如,同义词和上位词)。我们提出了将词向量转换为稀疏(可选二进制)向量的方法。结果表示更类似于NLP中通常使用的可解释特征,尽管它们是从原始语料库中自动发现的。因为向量是高度稀疏的,所以它们在计算上很容易处理。最重要的是,我们发现它们在基准任务上的表现优于原始向量。

英文摘要:

Current distributed representations of words show little resemblance to theories of lexical semantics. The former are dense and uninterpretable, the latter largely based on familiar, discrete classes (e.g., supersenses) and relations (e.g., synonymy and hypernymy). We propose methods that transform word vectors into sparse (and optionally binary) vectors. The resulting representations are more similar to the interpretable features typically used in NLP, though they are discovered automatically from raw corpora. Because the vectors are highly sparse, they are computationally easy to work with. Most importantly, we find that they outperform the original vectors on benchmark tasks.

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。

↑