Issue 2 - Best of NLP research, Bipolar Detection, OpenAI’s faster neural networks, Sarcastobot, …
Release Date: January 19, 2018
Summary of some of the latest and most important works on NLP (2016 - Current)
Source: 57 Summaries of Machine Learning and NLP Research
How emotion recognition through voice will help machines to have a healthy relationship with humans.
AI and Natural Language Processing used to improve care for heart failure patients.
On Code and Data…
OpenAI releases fast and memory-efficient deep neural networks via gradient-checkpointing.
Sarcastic conversational bot.
Bipolar Detection through Tweets. My lab is involved in this one! very cool stuff!
Source: MIT Technology Review
On Education and Research…
This recent paper proposes a transfer learning approach that is effective for various NLP tasks.
Source: Fine-tuned Language Models for Text Classification
A new book released by a Microsoft team talks about the societal issues of technologies such as AI.
Source: The Future Computed: Artificial Intelligence and its role in society
The Google Brain Team reviews the best of their ML and AI work for the year 2017.
Source: Google Research Blog
Are chatbots dead? This article discusses the current state of Facebook’s M (virtual assistant) and why Facebook is planning to shut it down.
How to capture the curiosity and imagination of your audience when delivering scientific presentations.
Source: The David Attenborough Style of Scientific Presentation
If you spot any errors or inaccuracies in this newsletter please open an issue.
Submit a pull request if you would like to add important NLP news here.