| Creator |
|
|
|
| Language |
|
| Publisher |
|
| Date |
|
| Source Title |
|
| Vol |
|
| Issue |
|
| Publication Type |
|
| Access Rights |
|
| Abstract |
国内で人気の高い動画投稿閲覧サービスにYouTube とニコニコ動画が有る. これらのサービスでは膨大な動画が蓄積されており，利用者が好みの動画を探すのは容易ではない. サービス運営側は検索語と並び替えによる動画検索を提示しているものの，細やかな動画検索は実現できていない. そこで，多数の視聴者が動画に付与するタグを用いた動画分類を考える. 本研究ではニコニコ動画を対象とし，動画分類や類似動画提示...のために，タグの出現頻度および共起出現を用いた．
It is very difficult to find favorite movie from recent poplar movie uploading and streaming service such as YouTube and Nicovideo, because there are a lot of movies in those sites. Movies are given not only metadata such as title, creator, upload date and movie length by the creator, but also annotations such as tags and comments by viewers. Those annotations may be used as folksonomy based data mining, because viewer may present his/her sympathy in annotation. In this paper, we analyzed hierarchy of tags using frequency of tag.show more
|
Export Mendeley
IPSJ-MPS81-nico