Natural Language Processing: A Machine Learning PerspectiveCambridge University Press, 2021年1月7日 - 484 頁 With a machine learning approach and less focus on linguistic details, this gentle introduction to natural language processing develops fundamental mathematical and deep learning models for NLP under a unified framework. NLP problems are systematically organised by their machine learning nature, including classification, sequence labelling, and sequence-to-sequence problems. Topics covered include statistical machine learning and deep learning models, text classification and structured prediction models, generative and discriminative models, supervised and unsupervised learning with latent variables, neural networks, and transition-based methods. Rich connections are drawn between concepts throughout the book, equipping students with the tools needed to establish a deep understanding of NLP solutions, adapt existing models, and confidently develop innovative models of their own. Featuring a host of examples, intuition, and end of chapter exercises, plus sample code available as an online resource, this textbook is an invaluable tool for the upper undergraduate and graduate student. |
內容
Introduction | 3 |
Counting Relative Frequencies | 25 |
Feature Vectors | 50 |
Discriminative Linear Classifiers | 73 |
A Perspective from Information Theory | 98 |
Hidden Variables | 120 |
Generative Sequence Labelling | 147 |
Discriminative Sequence Labelling | 169 |
Bayesian Network | 259 |
Neural Network | 289 |
Summary | 310 |
Neural Structured Prediction | 343 |
Working with Two Texts | 370 |
Pretraining and Transfer Learning | 396 |
Deep Latent Variable Models | 423 |
453 | |
常見字詞
according action addition algorithm attention calculated Chapter character classification compared Computational conditional consider consists constituent contains context corresponding count decoding defined denotes dependency derivation discussed distribution document embeddings encoding estimation event example expectation feature vector Figure Formally function given gives gradient head hidden input instance introduced Intuitively iteration label language models latent layer learning linear log-linear models loss LSTM machine maximising method model parameters namely neural network node obtain optimisation output pair parsing particular perceptron performance possible prediction probability problem random referred relations representation represents respectively result rule sample score segmentation semantic sentence separate sequence sequence labelling shown shows similar step structure Table tasks topic training data training objective transition tree update variables vector word