Building Natural Language Generation Systems

Building Natural Language Generation Systems

by Ehud Reiter, Robert Dale

ISBN-10: 0521620368

ISBN-13: 9780521620369

Pub. Date: 05/28/2003

Publisher: Cambridge University Press

This book explains how to build Natural Language Generation (NLG) systems - computer software systems which use techniques from artificial intelligence and computational linguistics to automatically generate understandable texts in English or other human languages, either in isolation or as part of multimedia documents, Web pages, and speech output systems. Typically…  See more details below


This book explains how to build Natural Language Generation (NLG) systems - computer software systems which use techniques from artificial intelligence and computational linguistics to automatically generate understandable texts in English or other human languages, either in isolation or as part of multimedia documents, Web pages, and speech output systems. Typically starting from some non-linguistic representation of information as input, NLG systems use knowledge about language and the application domain to automatically produce documents, reports, explanations, help messages, and other kinds of texts.

The book covers the algorithms and representations needed to perform the core tasks of document planning, microplanning, and surface realization, using a case study to show how these components fit together. It also discusses engineering issues such as system architecture, requirements analysis, and the integration of text generation into multimedia and speech output systems.

It is essential reading for researchers interested in NLP, Al, and HCI, and for developers interested in advanced document-creation technology.

Read More

Product Details

Cambridge University Press
Publication date:
Studies in Natural Language Processing Series
Product dimensions:
5.98(w) x 8.98(h) x 0.75(d)

Table of Contents

1.1The Research Perspective2
1.1.1Differences between NL Generation and NL Understanding2
1.1.2Sharing Knowledge between Generation and Understanding3
1.2The Applications Perspective4
1.2.1Computer as Authoring Aid5
1.2.2Computer as Author6
1.2.3Uses of NLG Technology6
1.3Some Example NLG Systems7
1.3.1Weather Reporter8
1.4A Short History of NLG19
1.5The Structure of This Book20
1.6Further Reading21
2National Language Generation in Practice23
2.2When Are NLG Techniques Appropriate?24
2.2.1Text versus Graphics25
2.2.2Natural Language Generation versus Mail Merge26
2.2.3Natural Language Generation versus Human Authoring28
2.3Using a Corpus to Determine User Requirements30
2.3.1Assembling an Initial Corpus of Output Texts31
2.3.2Analysing the Information Content of Corpus Texts33
2.3.3Creating a Target Text Corpus35
2.4Evaluating NLG Systems37
2.5Fielding NLG Systems38
2.6Further Reading40
3The Architecture of a Natural Language Generation System41
3.2The Inputs and Outputs of Natural Language Generation42
3.2.1Language as Goal-Driven Communication42
3.2.2The Inputs to Natural Language Generation43
3.2.3The Output of Natural Language Generation46
3.3An Informal Characterisation of the Architecture47
3.3.1An Overview of the Architecture47
3.3.2Content Determination50
3.3.3Document Structuring51
3.3.5Referring Expression Generation55
3.3.7Linguistic Realisation57
3.3.8Structure Realisation59
3.4The Architecture and Its Representations59
3.4.1Broad Structure and Terminology59
3.4.3The Document Planner63
3.4.4Document Plans64
3.4.6Text Specifications66
3.4.7Phrase Specifications67
3.4.8Surface Realisation71
3.5Other Architectures72
3.5.1Different Representations and Modularisations72
3.5.2Different Architectures: Integrated Systems76
3.6Further Reading77
4Document Planning79
4.1.1What Document Planning Is About79
4.1.2The Inputs and Outputs of Document Planning80
4.1.3A WeatherReporter Example82
4.2Representing Information in the Domain83
4.2.1What's in a Domain Model?86
4.2.2Domain Modeling for WeatherReporter87
4.2.3Implementing Domain Models88
4.2.4Defining Messages89
4.2.5Determining the Degree of Abstraction in Messages91
4.2.6A Methodology for Domain Modeling and Message Definition94
4.3Content Determination95
4.3.1Aspects of Content Determination96
4.3.2Deriving Content Determination Rules98
4.3.3Implementing Content Determination100
4.4Document Structuring101
4.4.1Discourse Relations102
4.4.2Implementation: Schemas104
4.4.3Implementation: Bottom-up Techniques107
4.4.4A Comparison of Approaches109
4.4.5Knowledge Acquisition110
4.5Document Planner Architecture110
4.6Further Reading112
5.1.1Why Do We Need Microplanning?115
5.1.2What's Involved in Microplanning?116
5.1.3The Inputs and Outputs of Microplanning117
5.1.4The Architecture of a Microplanner122
5.2.1Simple Lexicalisation124
5.2.2Simple Lexical Choice126
5.2.3Contextual and Pragmatic Influences on Lexical Choice128
5.2.4Expressing Discourse Relations129
5.2.5Fine-Grained Lexicalisation130
5.3.1Mechanisms for Sentence Formation133
5.3.2Choosing between Possible Aggregations140
5.3.3Order of Presentation142
5.3.4Paragraph Formation143
5.4Generating Referring Expressions144
5.4.1The Nature of the Problem144
5.4.2Forms of Referring Expressions and Their Uses145
5.4.3Requirements for Referring Expression Generation146
5.4.4Generating Pronouns149
5.4.5Generating Subsequent References152
5.5Limitations and Other Approaches156
5.6Further Reading157
6Surface Realisation159
6.2Realising Text Specifications162
6.3Varieties of Phrase Specifications164
6.3.1Skeletal Propositions165
6.3.2Meaning Specifications166
6.3.3Lexicalised Case Frames168
6.3.4Abstract Syntactic Structures168
6.3.5Canned Text169
6.3.6Orthographic Strings170
6.4.1An Overview171
6.4.2The Input to KPML171
6.4.3Using Systemic Grammar for Linguistic Realisation176
6.5.1An Overview179
6.5.2The Input to Surge180
6.5.3Functional Unification Grammar182
6.5.4Linguistic Realisation via Unification183
6.6.1An Overview186
6.6.2The Input to RealPro187
6.6.3Meaning-Text Theory189
6.6.4How RealPro Works190
6.7Choosing a Realiser192
6.8Bidirectional Grammars194
6.9Further Reading196
7Beyond Text Generation198
7.2.1The Uses of Typography201
7.2.2Typography in NLG Systems203
7.2.3Implementing Typographic Awareness206
7.3Integrating Text and Graphics208
7.3.1The Automatic Generation of Graphical Objects209
7.3.2Choosing a Medium210
7.3.3Commonalities between Text and Graphics213
7.3.4Implementing Text and Graphics Integration214
7.4.1Hypertext and Its Uses216
7.4.2Implementing Hypertext-based NLG Systems219
7.5Speech Output221
7.5.1The Benefits of Speech Output221
7.5.2Text-to-Speech Systems222
7.5.3Implementing Concept-to-Speech225
7.6Further Reading227
AppendixNLG Systems Mentioned in This Book229

Read More

Customer Reviews

Average Review:

Write a Review

and post it to your social network


Most Helpful Customer Reviews

See all customer reviews >