KDD19 Tutorial: From Shallow to Deep Language Representations: Pre-training, Fine-tuning, and Beyond
Location: TBATutors: Sheng Zha, Aston Zhang, Haibin Lin, Chenguang Wang, Mu Li, Alex J. Smola
Natural language processing (NLP) is at the core of the pursuit for artificial intelligence, with deep learning as the main powerhouse of recent advances. Most NLP problems remain unsolved. The compositional nature of language enables us to express complex ideas, but at the same time making it intractable to spoon-feed enough labels to the data-hungry algorithms for all situations. Recent progress on unsupervised language representation techniques brings new hope. In this hands-on tutorial, we walk through these techniques and see how NLP learning can be drastically improved based on pre-training and fine-tuning language representations on unlabelled text. Specifically, we consider shallow representations in word embeddings such as word2vec, fastText, and GloVe, and deep representations with attention mechanisms such as BERT. We demonstrate detailed procedures and best practices on how to pre-train such models and ne-tune them in downstream NLP tasks as diverse as finding synonyms and analogies, sentiment analysis, question answering, and machine translation. All the hands-on implementations are with Apache (incubating) MXNet and GluonNLP, and part of the implementations are available on Dive into Deep Learning.
We are targeting engineers, scientists, and instructors in the eld of natural language processing, data mining, text mining, deep learning, machine learning, and arti cial intelligence. While the audience with a good background in these areas would bene t most of this tutorial, it will give general audience and newcomers an introductory pointer to the presented materials.
|TBA||Basics of hands-on deep learning||ndarray, autograd|
|TBA||Shallow language representations in word embedding||word2vec, fasttext, GloVe, pre-train|
|TBA||Fine-tuning pre-trained word embedding||analogy, sa-rnn, sa-cnn|
|TBA||Attention mechanisms||attention, seq2seq with attention on mt|
|TBA||Deep language representations with attention|
|TBA||Fine-tuning pre-trained BERT||finetune|
|TBA||Q & A and Closing|