This book presents results from the "ESPRIT Project 393" "Construction and Interrogation of Knowledge Bases Using Natural Language Text and Graphics" (ACORD). The project involved more that 50 people over five years, and included industrial and academic partners from three countries. The main result of the project is a highly sophisticated human-machine interface allowing the construction and interrogation of a common knowledge base in English, French, and German, using both natural language and graphics. Chapter 1 presents an overview of the project. Chapters 2-4 develop specific issues concerning the grammar models used (functional grammars and unification categorical grammars). Chapter 5 discusses anaphora resolution and chapter 6 text generation. Chapter 7 explains the integration of text and graphics, and chapter 8 discusses the knowledge base module. The contributors are theoretically oriented and strongly concerned with generalization and the underlying background of potential further developments.
