Expressibility and the Problem of Efficient Text Planning
Published on 16. July 1992
Book
Hardback
224 pages
978-1-85567-022-8 (ISBN)
Description
Natural language generation is the process of turning a representation of a "situation" into an expression of some relevant portion of that situation in a natural language text. It is a field of artificial intelligence that is currently experiencing a rapid growth of interest among researchers. One of its central problems is that of "expressibility". The question is: How do we ensure that, when an Artificial Information system is planning the "message input", it does not put in something that is "expressible" in natural language? The problem is that, assuming a staged approach to generation, the constraints imposed by natural language itself are not available until the processing reaches the language stage. The author of this book presents her solution to this problem; the essence of which is a level of representation called a Text Structure, which is intermediate between the representation of the world and language itself.
More details
Series
Language
English
Place of publication
London
United Kingdom
Publishing group
Bloomsbury Publishing PLC
Target group
Professional and scholarly
Illustrations
figures, bibliograph
Dimensions
Height: 229 mm
Width: 152 mm
ISBN-13
978-1-85567-022-8 (9781855670228)
Copyright in bibliographic data is held by Nielsen Book Services Limited or its licensors: all rights reserved.
Schweitzer Classification
Other editions
Additional editions
E-Book
12/2015
1st Edition
Bloomsbury Academic
€155.99
Available for download
Content
Part 1 The generation gap: overview; defining the problem; Text Structure - an intermediate level of representation for text planning. Part 2 Terminology and issues: the generation process; issues. Part 3 Motivations from language and applications: analysis of revision; motivations from applications; summary. Part 4 The Text Structure: linguistic resources; determining the vocabulary of Text Structure; the Text Structure representation. Part 5 The architecture of Spokesman: overview of Spokesman; building the Text Structure; building the linguistic specification; control of the text planner. Part 6 Text Structure in action: incremental text generation from the "Main Street" simulation; Text Structure and portability. Part 7 Alternative architectures: traditional two components systems; non-traditional architectures; summary. Part 8 Conclusion: methodology; summary of the contributions; what does this work have to say to AI in general?; future excitements.