ht slidecreators @drlukesloan and dr @MattLWilliams and other contributors (see slide to the left which is the 1st of a presentation which is the source of much of this post's info)
You can find these and other slides - and so look at more detail behind the 1st 4 of this post's points at the national centre for research methods website.
You can find these and other slides - and so look at more detail behind the 1st 4 of this post's points at the national centre for research methods website.
Did you know that twitter might be?:
1) as good as traditional polls in predicting voting
patterns (conclusion from looking at sentiment in relation to candidates in German
general election - Tumasjan et al. – 2010)
2) as good at predicting movie revenue as the
Hollywood Stock Market (conclusion from correlating frequency and sentiment
related to movies on Twitter with their revenue - Asur & Huberman 2010)
3) more accurate at estimating the centre of
earthquakes than conventional methods (conclusion from analysis of twitter data
- Sakaki et al. - 2010)
4) as good as the census data at identifying crime
hotspots (Mallerson and Andersen 2014)
5) as good as and better than all of the above in using its (your?) data to find out about you ( @uinp 2014 )
No comments:
Post a Comment