
|
Department of
Computational
Perception
|

|
Special Topics: Social Media Mining and
Analysis
344.053,
KV, 3h (4.5 ETCS), Winter Term 2017/18
Assoc.-Prof.
Dr. Markus Schedl
Time: Monday, 13:45 -
15:15
Start: October 2, 2017
Location: S3 055
Description
This course will be held foremost as an exercise in which students will
implement and refine existing approaches to a task in context of social
media mining or analysis. The work will be conducted in teams, each of
which will select a topic to work on. The requirements to complete the
course (and earn a positive grade) are the following:
- gaining an understanding of the selected task and existing solutions
- conducting a literature research and writing a short report about existing work on the topic (max. 3 pages)
- implementing and evaluating a refined version of an existing approach (or an innovate novel approach)
- giving a presentation of the topic, implemented approach/solution, and its evaluation (about 30 minutes per team)
- writing a scientific article about the approach in the style of a conference submission (about 10 pages)
Topics students can choose from include, but are not limited to the following (own suggestions are welcome too):
- influential user detection
- network structure analysis
- sentiment analysis
- flow analysis
- analyzing aspects such as similarity or diversity in user preferences
- categorization of users from their digital traces (demographics, mood, personality, habbits, etc.)
- studying
the interrelationship between music taste and consumption patterns
(e.g., time of day) or other user characteristics (e.g., country)
- trend detection and popularity prediction
- localizing microblog messages or other multimedia material
- real-world event detection
- country of origin detection (for persons, music artists, bands, or other items)
Many other possible topics can be derived, for instance, from the proceedings of the ICWSM conference:
http://icwsm.org/2016/program/program
http://icwsm.org/2015/program/program
http://icwsm.org/2014/program/program
...
Also the following book gives an overview about basics in mining user-generated content:
Marie-Francine
Moens, Juanzi Li, Tat-Seng Chua (eds.), Mining of User Generated
Content and Its Applications, CRC Press, January 2014.
Schedule
The detailed schedule can be found on the course's KUSSS
web page.
last edited by ms 2017-09-04