2017 Fiscal Year Final Research Report
Information Compilation Technologies for Comic Contents
| Project/Area Number |
15K12103
|
|---|
| Research Category |
Grant-in-Aid for Challenging Exploratory Research
|
| Allocation Type | Multi-year Fund |
|---|
| Research Field |
Intelligent informatics
|
|---|
| Research Institution | Kansai University |
|---|
Principal Investigator |
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
山西 良典 立命館大学, 情報理工学部, 助教 (50700522)
|
|---|
| Co-Investigator(Renkei-kenkyūsha) |
NAGAMORI Mitsuharu 筑波大学, 図書館情報メディア系, 講師 (60272209)
|
|---|
| Research Collaborator |
YAMASHITA Ryo 関西大学, 大学院・総合情報学研究科
PARK Byeongseon 関西大学, 大学院・総合情報学研究科
AKAI Yuki 関西大学, 大学院・総合情報学研究科
|
|---|
| Project Period (FY) |
2015-04-01 – 2018-03-31
|
| Keywords | コミック工学 / 探索的検索 / 情報編纂 |
|---|
| Outline of Final Research Achievements |
Since comic is cross-modal content in which pictures and letters are used complementarily and cooperatively, it is necessary to store the contents of the comic in a structured manner in advance to make these contents available on the computer. In this research, information on digitalized comics was systematically encoded and hierarchically annotated in RDF format to meet the requirement. By referring to these annotation sets, the computer can search and process comic contents along with the user's intention.
In addition, focusing on review sentences as clues to acquire the contents of comics, we proposed a method of acquiring topics by combining term frequency - inverse document frequency (tf-idf) method and hierarchical Latent Dirichlet Allocation (hLDA) method to classify comics. Using these data and methods, a comic question answering system and an exploratory comic search system are developed.
|
| Free Research Field |
インタラクションデザイン
|
