DocumentCode
1793614
Title
Estimation of inheritance relationship between contents on social media — Case study of Niconico as a video-sharing site
Author
Ohta, Hitoyoshi ; Kobayashi, Akihiro ; Masuyama, Shigeru
Author_Institution
Comput. Sci. & Eng., Toyohashi Univ. of Technol., Toyohashi, Japan
fYear
2014
fDate
20-21 Aug. 2014
Firstpage
191
Lastpage
196
Abstract
Social media is a platform on which collaborative creative activities on the Web create new contents. This study proposes a method for estimating inheritance relationship between contents when we assumed that contents have inherited from some other contents in its parts or properties. We estimate this relationship by targeting Niconico, a typical video-sharing Website in Japan, because we found this relationship from various meta-data given to the video. We think that we can observe the culture formation that occurs in social media by accounting for inheritance relationship of contents. We also think that we can analyze the phenomenon that a piece of contents inspired another one in chain in social media by applying the visualized culture formation. First, we propose a reference graph that represents inheritance relation obtained by analyzing the description of videos. We investigate characteristic of the relationship using edges in this graph. We confirm that some contents in Niconico has inheritance relationship with other contents. We have found that relationship is classified into the following four types: (1) Remake: Creator remakes some other videos into high-quality sound or high definition. (2) Arrangement: Creator arranges some other videos in their favorite style. (3) Partially-used: Creator uses some other videos in its music or scene as materials. (4) Inspired: Creator refers to some other videos in its representation or technique. We also propose a method for estimating this relationship using tags and titles from videos. We conducted experiments to evaluate our method and attained F-value 0.444 when threshold value is 0.330.
Keywords
Internet; meta data; pattern classification; social networking (online); F-value; Niconico; arrangement type; classification; inheritance relationship estimation; inspired type; meta-data; partially-used type; reference graph; remake type; social media; video-sharing Website; Classification algorithms; Educational institutions; Games; Informatics; Media; Vectors; Video sharing; Web mining;
fLanguage
English
Publisher
ieee
Conference_Titel
Advanced Informatics: Concept, Theory and Application (ICAICTA), 2014 International Conference of
Conference_Location
Bandung
Print_ISBN
978-1-4799-6984-5
Type
conf
DOI
10.1109/ICAICTA.2014.7005939
Filename
7005939
Link To Document