Paper Reading: Papers in Frontiers of NLP 2018 collection

1.Papers collections
Note: the original name of the paper will be appended soonly!

Index	Paper	Year	Brief Intro	Note
1.	[Collobert & Weston, ICML ’08]	2008	Multi-task learning.	MTL: Win Test-of-time-award at ICML 2018
2.	[Pennington et al., EMNLP ’14; Levy et al., NIPS ’14]	2014	Generate embeeidng by matrix factorization	New method of embedding
3.	[Levy et al., TACL ’15]	2015	Classic methods (eg. PMI and SVD) for embedding generation	New method of embedding
4.	[Le & Mikolov, ICML ’14; Kiros et al., NIPS ’15]	2016	Skip-gram for sentence representation	Skip-gram
5.	[Grover & Leskovec, KDD ’16]	2016	Skip-gram for Nueral Network modelling	Skip-gram
6.	[Luong et al., ’15]	2015	Difference embedding projection aids trasfer learning	Embedding projection
7.	[Hochreiter & Schmidhuber, NeuComp ’97]	1997	The original paper for LSTM	LSTM
8.	[Kalchbrenner et al., ’17]	2017	Dilated CNN	CNN: To enable wider receptive field
9.	[Wang et al., ACL ’16]	2016	Stacked LSTM and CNN	Stacked model
10.	[Bradbury et al., ICLR ’17]	2017	Use convolution to speed up LSTM	CNN&LSTM combination
11.	[Tai et al., ACL ’15]	2015	Extend Recursive nueral netword to LSTM	Recursive neural network put forward
12.	[Bastings et al., EMNLP ’17]	2017	graph convolutional neural network	Cnn over graph(trees)
13.	[Levy and Goldberg, ACL ’14]	2014	word embeddings generated form dependencies	Embedding generation
14.	[Wu et al., ’16]	2016	Deep LSTM	New seq2seq model
15.	[Kalchbrenner et al., arXiv ’16; Gehring et al., arXiv ’17]	2017	Convolutional encoders	New seq2seq model
16.	[Vaswani et al., NIPS ’17]	2017	Transformer: pure attention architecture	New seq2seq model
17.	[Chen et al., ACL ’18]	2018	combination of LSTM and Transformer	New seq2seq model
18.	[Vinyals et al., NIPS ’16]	2016	Attention in one-shot learning	Attention & one-shot
19.0	[Graves et al., arXiv ’14]	2014	Neural Turing Machine	Memory Network
19.1	[Weston et al., ICLR ’15]	2015	Memory Network	Memory Network
19.2	[Sukhbaatar et al., NIPS ’15]	2015	End-to-end Memory Networks	Memory Network
19.3	Dynamic Memory Networks [Kumar et al., ICML ’16]	2016	Dynamic Memory Networks	Memory Network
19.4	[Graves et al., Nature ’16]	2016	Neural Differentiable Computer	Memory Network
19.5	[Henaff et al., ICLR ’17]	2017	Recurrent Entity Network	Memory Network
20.	[Peters et al., NAACL ’18],之前看过一篇稍后补上	2018	Language model embedding used as feature	Language model
21.	[Howard & Ruder, ACL ’18]	2018	Language model fine tuned on task data	Language model
22.	[Jia & Liang, EMNLP ’17]	2017	Adversarial examples	Adversarial
23.	[Miyato et al., ICLR ’17; Yasunaga et al., NAACL’18]	2018	Adversarial training	Form of regularization
24.	[Ganin et al., JMLR ’16; Kim et al., ACL ’17]	2017	Domain adversarial loss	Form of regularization
25.	[Semeniuta et al., ’18]	2018	GANs’ application in NLG	GAN for NLP
26.	[Paulus et al., ICLR ’18]	2018	RL for summarization	RL with ROUGE loss
27.	[Ranzato et al., ICLR ’16]	2016	RL for Machine Translation	RL with BLUE loss
28.	[Conneau et al., ICLR’18]	2018	word translation without parallel data	Low-resource scenarios