2013 Fiscal Year Final Research Report
Realization of Discussion Mining Using Music Theory for Meeting Record Analysis
| Project/Area Number |
23500145
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Research Field |
Media informatics/Database
|
|---|
| Research Institution | Future University-Hakodate |
|---|
Principal Investigator |
HIRATA Keiji 公立はこだて未来大学, システム情報科学部, 教授 (30396121)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
TOJO Satoshi 北陸先端科学技術大学院大学, 情報科学研究科, 教授 (90272989)
HAMANAKA Masatoshi 筑波大学, システム情報工学研究科(系), 講師 (30451686)
NAGAO Katashi 名古屋大学, 情報科学研究科, 教授 (70343209)
|
|---|
| Project Period (FY) |
2011 – 2013
|
| Keywords | コンテンツ / 音楽分析 / 意味構造解析 / 再利用 |
|---|
| Research Abstract |
There are the following three results. (1) We have theoretically investigated the difference between the events included in meeting records and musical events. We have succeeded in formalizing time-span trees and building a computational model. (2) As a milestone toward a discussion structure analyzer for a discussion mining system, we have progressed to improve and enhance the GTTM analyzer to complete enabling to handle a polyphony. (3) We have verified the developed method to efficiently extract the semantic structure of discussion. We have reconfigurated the GTTM rules in the thinking way of GTTM, and we succeeded automatic generation of a discussion time-span tree from multimedia discussion records.
|
