Manning Publications, 2019. — 327 p. — ISBN: 1617294799.
Deep Learning for Search is a practical book about how to use (deep) neural networks to help build effective search engines. This book examines several components of a search engine, providing insights on how they work and guidance on how neural networks can be used in each context. Emphasis is given to practical, example-driven explanations of search and deep learning techniques, most of which are accompanied by code. At the same time, references to relevant research papers are provided where applicable to encourage you to read more and deepen your knowledge on specific topics. Neural network and search-specific topics are explained throughout the book as you read about them. After reading this book, you’ll have a solid understanding of the main challenges related to search engines, how they are commonly addressed, and what deep learning can do to help. You’ll gain a solid understanding of several different deep learning techniques and where they fit in the context of search. You’ll get to know the Lucene and Deeplearning4j libraries well. In addition, you’ll develop a practical attitude toward testing the effectiveness of neural networks (rather than viewing them as magic) and measuring their costs and benefits.
This book is intended for readers with an intermediate programming background. It will be best if you’re proficient in Java programming, with an interest or active involvement in developing search engines. You should read this book if you would like to make your search engine more effective at giving relevant results and therefore more useful for end users. Even if you don’t have a search background, basic concepts about search engines are introduced during the course of the book as each specific aspect of search is touched on. Similarly, you aren’t expected to already know about machine or deep learning. This book will introduce all the required machine learning and deep learning basics, together with practical tips regarding the application of deep learning to search engines in production scenarios. You should be ready to get your hands on the code and extend existing open-source libraries to implement deep learning algorithms to solve search problems.
About this bookAbout the cover illustration.
Search meets deep learning
Neural search.
Generating synonyms.
Throwing neural nets at a search engine
From plain retrieval to text generation.
More-sensitive query suggestions.
Ranking search results with word embeddings.
Document embeddings for rankings and recommendations.
One step beyond
Searching across languages.
Content-based image search.
A peek at performance