Applying Large Language Models to Search
摘要
This chapter introduces the reader to ideas, research and practical industry-level applications that use LLMs to improve the user experience of finding the information they need. Encoding the meaning of the text in a numeric vector format, being able to process input/output in natural language, expanding and rephrasing documents and queries to improve their “match-ability” and supporting search across media and language are just the tip of the iceberg. This chapter guides the reader through the state-of-the-art and industry adoptions, highlighting the positive and not-so-positive aspects to give an unbiased and comprehensive perspective.