Skip to content

I will analyse sentiment in tweets concerning the German parties.

Notifications You must be signed in to change notification settings

FranziSchmitt/Twitter_sentimentanalysis

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

74 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Twitter topic analysis

This is the portforlio project, I worked on while I participated in Data Science Retreat in Berlin.

I set up a Raspberry Pi and scraped for Tweets that mention at least one of the German parties.

In a first attempt to find patterns, I tried k-means clustering of the tweets. However, using Latent Dirichlet Allocation to find topics proved to be the better way to go. The quality of the topics heavily depends on the quality of the preprocessing. Since tweets, in all their shortness, contain heaps of noise, the right amount of text preprocessing was key.

In addition to finding topics, I also looked at topics over time and could see that topic popularity fits political events taking place in Germany.

This is still an ongoing project. Therefore, the code is not in a final format.

About

I will analyse sentiment in tweets concerning the German parties.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published