Hyperbolic Attention Networks, Whistle to Music, Graph2Seq, Netflix Research, Intel NLP Tools,…
Welcome to the 16th Issue of the NLP Newsletter! Here is this week’s notable NLP news: Philosophical view of word embeddings; Netflix research website; Intel NLP python tools; DeepMind explores dopamine’s role in brain; Emotion flow in books; whistle to music; Hyperbolic attention networks; Graph to sequences; and much more…
On People…
A philosophical view of word embeddings and how they are deeply and linguistically motivated — Link
Modeling the flow of emotions over a book using RNNs (published in NAACL 2018) — Link
AI is harder than you think (opinion piece by Gary Marcus and Ernest Davis) — Link
Facebook AI researchers turn whistles into orchestral music using style transfer — Link
How DeepMind is using AI to explore dopamine’s role in learning — Link
On Education and Research…
How neural models, like LSTM, use context — Link
Hyperbolic attention networks that can match the complexity of data with hierarchical and power-law structure (useful for generalizing neural machine translation tasks) — Link
A step-by-step guide on how to solve 90% of NLP problems — Link
Deep exploration on how deep learning architectures model the compositionality of text sequences — Link
A list of the best free NLP and Machine Learning educational resources — Link
An interesting paper on teaching machines the ability to learn textual/visual relationships — Link
Graph2Seq models graph using graph-based neural network through novel aggregation strategy mechanism — Link
On Code and Data…
1000+ games for reinforcement learning research by OpenAI — Link
A dataset and model for comprehending paragraphs describing processes — Link
How to obtain a good corpus for grammar correction via translation — Link
On Industry…
The race for open source neural machine translation — Link
Netflix announces research website on Medium (discuss research and ML resources) — Link
Hugging Face, NLP-based startup, raises $4M to create emotional intelligent chatbots — Link
Intel open source python library for exploring state-of-the-art deep learning techniques for NLP — Link
Quotes of the week…
On context for understanding linguistic phenomena
Illustrations of the week…
Opportunities in linguistics — Source
Worthy Mentions…
Understanding LSTM and its diagrams — Link
The role of AI and NLP in customer experience — Link
If you spot any errors or inaccuracies in this newsletter please comment below. I discuss more about NLP and Deep Learning on Twitter.