● Machine Comprehension / Question Anxswering
-Memory Module
Neural Turing Machines arxiv
End-To-End Memory Networks arxiv
Ask Me Anything: Dynamic Memory Networks for Natural Language Processing arxiv
Dynamic Memory Networks for Visual and Textual Question Answering arxiv
Tracking the World State with Recurrent Entity Networks arxiv
Learning to Skim Text arxiv
-Embedding Module
A Comparative Study of Word Embeddings for Reading Comprehension arxiv
A Structured Self-attentive Sentence Embedding arxiv
-TOEFL
Towards Machine Comprehension of Spoken Content: Initial TOEFL Listening Comprehension Test by Machine arxiv
Hierarchical Attention Model for Improved Machine Comprehension of Spoken Content arxiv
-Representation of words
Exploiting Similarities among Languages for Machine Translation arxiv
Distributed Representations of Sentences and Documents arxiv
Skip-Thought Vectors arxiv
Learning Context-Specific Word/Character Embeddings [AAAI] (https://aaai.org/ocs/index.php/AAAI/AAAI17/paper/view/14601/14266) [2017/2] ├ Learn Multi-Sense Vector for different meannings of single word, letting machine to deicide when to generate new sense
Learned in Translation: Contextualized Word Vectorsarxiv [2017/8] ├ Learn Word vectors by training on English-German Translation(ATTENTIONAL SeqtoSeq)
● Representation of Audio
Audio Word2Vec arxiv [2016/3]
Deep convolutional acoustic word embeddings using word-pair side information arxiv [2015/10]
Learning Latent Representations for Speech Generation and Transformation arxiv [2017/4]
● Text Generation
Controllable Text Generation arxiv [2017/3]
● Summary
Get To The Point: Summarization with Pointer-Generator Networks https://arxiv.org/pdf/1704.04368.pdf
ABSTRACTIVE HEADLINE GENERATION FOR SPOKEN CONTENT BY ATTENTIVE RECURRENT NEURAL NETWORKS WITH ASR ERROR MODELING https://arxiv.org/pdf/1612.08375.pdf
Order-Preserving Abstractive Summarization for Spoken Content Based on Connectionist Temporal Classification https://arxiv.org/pdf/1709.05475v2.pdf
A Deep Reinforced Model for Abstractive Summarization https://arxiv.org/abs/1705.04304
LEARNING TO ENCODE TEXT AS HUMAN-READABLE SUMMARIES USING GENERATIVE ADVERSARIAL NETWORKS https://openreview.net/pdf?id=r1kNDlbCb
Diversity driven Attention Model for Query-based Abstractive Summarization arxiv [2017/4] ├ Solve the problem of redundant, repitetive words in summary. Use math-orthogonal to obtain attention vector.
● Generalized / Miscellaneous
Attention Is All You Need [arxiv] (https://arxiv.org/abs/1706.03762) [2017/6] ├ No Convolution and recurrence, pure attention. Use multi-head to grasp various attention of the inputs ( self-attention )