(42) Experimental System for Argument Text Generation: Dulcinea
Machine: Multi-PSI or pseudo-PIM/m
Environment: PIMOS
Language: KL1
Source Code: 2MB
Documents: Manual (English/Japanese)
Overview
Dulcinea generates Japanese text that justifies a given goal.
Configuration
The system consists of following four modules.
- Semantic Contents Generator:
- This module creates an Argument Graph which represents the
semantic content of arguments.
- FTS Generator:
- This module creates an FTS(Functional Text Structure) from a
given argument graph.
- Text Structure Generator:
- This module adds order and connection information to the FTS.
- Realizer:
- Tense, aspect and mood are fixed in this module.
Function
- Beliefs consist of facts,rules and judgments for the state of
affairs. The system makes arguments in various domains by exchanging
the contents of a belief database.
- This system can generate not only direct grounds,but also
refutations for the expected opposing arguments,together with examples
that are indispensable to persuasive argument.
- By varying the argument strategies and criteria for sentence
structure,you can perform the text generation experiment and conduct
research into the objective criteria for natural text.
Reference
- IKEDA Teruo,KOTANI,Akira,HAGIWARA Kaoru and KUBO Yukihiro
``Argument Text Generation System(Dulcinea),'' In the Proceedings
of International Conference on Fifth Generation Computer Systems
1992
FTP
- Experimental System for Argument Text Generation: Dulcinea
[873K]
www-admin@icot.or.jp