Neural machine translation
Philipp Koehn.
Bok Engelsk 2020 · Electronic books.
| Omfang | 1 Recurso online
|
|---|---|
| Opplysninger | Title from publisher's bibliographic system (viewed on 01 Jun 2020).. - Cover -- Half-title -- Title page -- Copyright information -- Dedication -- Table of contents -- Preface -- Reading Guide -- Part I Introduction -- 1 The Translation Problem -- 1.1 Goals of Translation -- 1.2 Ambiguity -- 1.3 The Linguistic View -- 1.4 The Data View -- 1.5 Practical Issues -- 2 Uses of Machine Translation -- 2.1 Information Access -- 2.2 Aiding Human Translators -- 2.3 Communication -- 2.4 Natural Language Processing Pipelines -- 2.5 Multimodal Machine Translation -- 3 History -- 3.1 Neural Networks -- 3.2 Machine Translation -- 4 Evaluation -- 4.1 Task-Based Evaluation -- 4.2 Human Assessments -- 4.3 Automatic Metrics -- 4.4 Metrics Research -- Part II Basics -- 5 Neural Networks -- 5.1 Linear Models -- 5.2 Multiple Layers -- 5.3 Nonlinearity -- 5.4 Inference -- 5.5 Back-Propagation Training -- 5.6 Exploiting Parallel Processing -- 5.7 Hands On: Neural Networks in Python -- 6 Computation Graphs -- 6.1 Neural Networks as Computation Graphs -- 6.2 Gradient Computations -- 6.3 Hands On: Deep Learning Frameworks -- 7 Neural Language Models -- 7.1 Feed-Forward Neural Language Models -- 7.2 Word Embeddings -- 7.3 Noise Contrastive Estimation -- 7.4 Recurrent Neural Language Models -- 7.5 Long Short-Term Memory Models -- 7.6 Gated Recurrent Units -- 7.7 Deep Models -- 7.8 Hands On: Neural Language Models in PyTorch -- 7.9 Further Readings -- 8 Neural Translation Models -- 8.1 Encoder-Decoder Approach -- 8.2 Adding an Alignment Model -- 8.3 Training -- 8.4 Deep Models -- 8.5 Hands On: Neural Translation Models in PyTorch -- 8.6 Further Readings -- 9 Decoding -- 9.1 Beam Search -- 9.2 Ensemble Decoding -- 9.3 Reranking -- 9.4 Optimizing Decoding -- 9.5 Directing Decoding -- 9.6 Hands On: Decoding in Python -- 9.7 Further Readings -- Part III Refinements -- 10 Machine Learning Tricks -- 10.1 Failures in Machine Learning.. - 10.2 Ensuring Randomness -- 10.3 Adjusting the Learning Rate -- 10.4 Avoiding Local Optima -- 10.5 Addressing Vanishing and Exploding Gradients -- 10.6 Sentence-Level Optimization -- 10.7 Further Readings -- 11 Alternate Architectures -- 11.1 Components of Neural Networks -- 11.2 Attention Models -- 11.3 Convolutional Machine Translation -- 11.4 Convolutional Neural Networks with Attention -- 11.5 Self-Attention: Transformer -- 11.6 Further Readings -- 12 Revisiting Words -- 12.1 Word Embeddings -- 12.2 Multilingual Word Embeddings -- 12.3 Large Vocabularies -- 12.4 Character-Based Models -- 12.5 Further Readings -- 13 Adaptation -- 13.1 Domains -- 13.2 Mixture Models -- 13.3 Subsampling -- 13.4 Fine-Tuning -- 13.5 Further Readings -- 14 Beyond Parallel Corpora -- 14.1 Using Monolingual Data -- 14.2 Multiple Language Pairs -- 14.3 Training on Related Tasks -- 14.4 Further Readings -- 15 Linguistic Structure -- 15.1 Guided Alignment Training -- 15.2 Modeling Coverage -- 15.3 Adding Linguistic Annotation -- 15.4 Further Readings -- 16 Current Challenges -- 16.1 Domain Mismatch -- 16.2 Amount of Training Data -- 16.3 Rare Words -- 16.4 Noisy Data -- 16.5 Beam Search -- 16.6 Word Alignment -- 16.7 Further Readings -- 17 Analysis and Visualization -- 17.1 Error Analysis -- 17.2 Visualization -- 17.3 Probing Representations -- 17.4 Identifying Neurons -- 17.5 Tracing Decisions Back to Inputs -- 17.6 Further Readings -- Bibliography -- Author Index -- Index.. - Deep learning is revolutionizing how machine translation systems are built today. This book introduces the challenge of machine translation and evaluation - including historical, linguistic, and applied context -- then develops the core deep learning methods used for natural language applications. Code examples in Python give readers a hands-on blueprint for understanding and implementing their own machine translation systems. The book also provides extensive coverage of machine learning tricks, issues involved in handling various forms of data, model enhancements, and current challenges and methods for analysis and visualization. Summaries of the current research in the field make this a state-of-the-art textbook for undergraduate and graduate classes, as well as an essential reference for researchers and developers interested in other applications of neural methods in the broader field of human language processing.
|
| Emner | |
| Sjanger | |
| Dewey | |
| ISBN | 9781108608480
|
