Sparse Overcomplete Word Vector Representations
作者： Manaal Faruqui , Yulia Tsvetkov , Dani Yogatama , Chris Dyer , Noah A. Smith
DOI: 10.3115/V1/P15-1144
关键词:
摘要: Current distributed representations of words show little resemblance to theories lexical semantics. The former are dense and uninterpretable, the latter largely based on familiar, discrete classes (e.g., supersenses) relations synonymy hypernymy). We propose methods that transform word vectors into sparse (and optionally binary) vectors. resulting more similar interpretable features typically used in NLP, though they discovered automatically from raw corpora. Because highly sparse, computationally easy work with. Most importantly, we find outperform original benchmark tasks.
arxiv.org
aclweb.org参考文章(58)
Andre Martins, Mario Figueiredo, Pedro Aguiar, Noah Smith, Structured Sparsity in Structured Prediction empirical methods in natural language processing. pp. 1500- 1511 ,(2011)
Yoshua Bengio, Xavier Glorot, Antoine Bordes, Antoine Bordes, Domain Adaptation for Large-Scale Sentiment Classification: A Deep Learning Approach international conference on machine learning. pp. 513- 520 ,(2011)
Manaal Faruqui, Chris Dyer, Improving Vector Space Word Representations Using Multilingual Correlation conference of the european chapter of the association for computational linguistics. pp. 462- 471 ,(2014) , 10.3115/V1/E14-1049
Brian Murphy, Tom Mitchell, Partha Talukdar, Learning Effective and Interpretable Semantic Models using Non-Negative Sparse Embedding international conference on computational linguistics. pp. 1933- 1950 ,(2012)
Rafal Zdunek, Shun-ichi Amari, Andrzej Cichocki, Anh Huy Phan, Nonnegative Matrix and Tensor Factorizations: Applications to Exploratory Multi-way Data Analysis and Blind Source Separation ,(2009)
Eric P. Xing, Jun Zhu, Sparse Topical Coding arXiv: Learning. ,(2012)
Alex Teichman, Andrew Y. Ng, Honglak Lee, Rajat Raina, Exponential family sparse coding with applications to self-taught learning international joint conference on artificial intelligence. pp. 1113- 1119 ,(2009)
Manaal Faruqui, Jesse Dodge, Sujay Kumar Jauhar, Chris Dyer, Eduard Hovy, Noah A. Smith, Retrofitting Word Vectors to Semantic Lexicons north american chapter of the association for computational linguistics. pp. 1606- 1615 ,(2015) , 10.3115/V1/N15-1184
Joshua T. Goodman, Exponential priors for maximum entropy models north american chapter of the association for computational linguistics. pp. 305- 312 ,(2005)
Peter D. Turney, Mining the web for synonyms: PMI-IR versus LSA on TOEFL european conference on machine learning. pp. 491- 502 ,(2001) , 10.1007/3-540-44795-4_42