From Shakespeare to Trump: AI that Talks with Attitude by Cheuk Ting Ho

Abstract:

One main application of AI (artificial intelligence) is in NLP (natural language processing) which allows the machine to understand neutral languages and learn a lot from the context of texts. With tools like Keras (using Tensorflow as backend) we can build a neutral network with only 3 layers (Embedding, LSTM, Softmax output) in order to "teach" a machine to generate articles, plays or speeches in the style of the training corpus and have lots of fun. Will it be like a Shakespearean actor in one instance and talk like Trump in another? How well is the AI against human in writings now and in the future? How creative could an AI be?

About Cheuk:

With a research background in theoretical physics, Cheuk has transferred her analytical and logical skills in natural science and built a career in Data Science. Cheuk is now a Data Scientist which has broad experience with time series forecasting including ARIMA and Holt-Winters models and machine learning like neural network and deep learning. Interested in NLP, Cheuk is also familiar with RNN, Attention model using bidirectional LSTM, GRU etc. Cheuk has also actively contributed to open source projects like Pandas, Gensim and Dateutil. Believing in gender equality, Cheuk is currently a co-organizer of AI club for Gender Minorities to support Tech Diversity and Inclusion.

Twitter: @cheukting_ho

Linkedin: https://www.linkedin.com/in/cheukting-ho/