2004 Fiscal Year Final Research Report Summary
Computational and Logical Analysis of Pressuposition
| Project/Area Number |
14580406
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Intelligent informatics
|
|---|
| Research Institution | The University of Tokyo |
|---|
Principal Investigator |
NAKAZAWA Tsuneko The University of Tokyo, Graduate School of Arts and Science, Associate Professor, 大学院・総合文化研究科, 助教授 (00292839)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
KATO Tsuneaki The University of Tokyo, Graduate School of Arts and Science, Associate Professor, 大学院・総合文化研究科, 助教授 (60334299)
HATAKEYAMA Shinichi The University of Tokyo, Graduate School of Arts and Science, Assistant Professor, 大学院・総合文化研究科, 助手 (20361587)
|
|---|
| Project Period (FY) |
2002 – 2004
|
| Keywords | presupposition / formal semantics / pragmatics / inference / abduction |
|---|
| Research Abstract |
It is well known that, in addition to truth-conditional meanings, dialogues give rise to secondary meaning known as presupposition which is derived from sentences contained in the dialogue. Presuppositions constitute a condition for a sentence to be assigned a truth value. For example, the sentence, "Taro regrets that he has studied cognitive science," is assigned a truth-value only because it presupposes that Taro has studied cognitive science. It cannot be always assumed that presuppositions are explicitly expressed by previous sentences in the dialogue, and therefore, it is necessary to calculate presupposed meanings based upon the preceding dialogue. Furthermore, presuppositions are realized, i.e. projected, in more than one way. In this research :
(1)To isolate the way presuppositions are projected from preceding sentences, linguistic data have been collected and analyzed.
(2)Based upon the data, the inference process to calculate the projected presuppositions has been formalized. The formalization is cast in the framework of Satisfaction Theory and Anaphoric Resolution, and the predictions made by the model has been evaluated.
(3)The projection model has been embedded in a computer system for natural language processing, as part of its context analyzer.
|
