Skip to content
Snippets Groups Projects

Add paper LeaveAComment

Merged Jan Lukas Steimann requested to merge LeaveACommentPaper into master
1 file
+ 51
0
Compare changes
  • Side-by-side
  • Inline
+ 51
0
# Zusammenfassung des Papers Leave a comment! A in-depth analysis of user comments on youtube
* User Kommentare sind die häufigste und kontroverseste Form der Kommunikation auf YouTube
* Dennoch werden jeden Tag werden große Mengen an Kommentaren gepostet
* Die Autoren des Papers stellen eine Erklärung für diese Gegensatz vor basiernd auf einem neuen Klassifkationsansatz für Kommentare
## Introduction
* Online videos are an object of social exchange
* Authors conduct a survey on the perception of comments
* Rather negative view on comments
* The way comments are presented is substandard
* Sequential list of comments sorted by creation date
* Only first two or three comments can be viewed on orginal visiable space
* => Places one and two in comment section are privileged
## Research Model and Data Collection
* First authors hat to find out how users communicate via comments
* User comments are basically many unstructured text fragments
* Authos focused on three basic comment types:
* T1: **Discussion Post** contains comments which are part of a discussion among users
* T2: **Inferior Comment** contains offensive statements and/or insults
* T3: **Substantial Comment** Conatins comment without offensive statement that carry certain content information
* Authors propose that T1 and T3 comments provide added value for users (Information, Entertainment, Social Exchange)
## Consistency Check: Is our Operationalization valid?
* Authors examined distribution of comment classes in each video category
### Impications
* Results indicate that no comment type is dominant on YouTube
* user communicate on differnt video topics in different ways
* Two comment types offer added value to user
* Main reason for comments
* Negative impressions come from comment type T2
## Comment Types and Rating - is there a relationship
* Author analyse how comments relate to the dispersion of likes and dislikes
* Two finding:
* Number of T3 comments have the strongest effect on relative amount of likes
* T2-comments have an influence on amount of dislikes
## Discussion
* Analysis yield several recommendations for improving user acceptance of comments
* Appropriate visualisation which take into context dependencies for particular video sequences into account could highlite valuable posts
* Dynamic and media based annotations are likely to be very well suited for visualisation or context-specific user comments
* A secondary rating system which user could use to express their emotional attitude
* Short emotional posts would retain their significance if a more suitable visualisation form was available
Loading