Archives for : Social Media

Data Statistics View

Today is my last day at my Project Assistent job at Vienna University of Technology. I did some summing up of my work and polished the TwitterAPI and also the Data Statistics View code. I want to share my implemnetation of the Data Statistics View code. This was done with html, php, javascript and SQL.  The project can be used for any data types stored in a SQL database.

One of my tasks at university was to download data from the Twitter public stream and analyse it. This work was easier with a tool that allows visualizing the number of downloads per hour/day/month.
The API I used to download tweets  is the one based on Adam Green’s implementation called 140dev. He also has a visualizing tool for the downloaded tweets. However this has less to do with numbers rather much more with the tweet texts.

The code for my implementation can be found on my GitHub repository.
It contains simple bar charts of the number of tweets downloaded.

Bar chart example
 

Working with the Twitter public stream I navigated a lot of questions which I found or did not find answeres to:

  • How can one download tweets only for a specific country?
  • When is the rate limit reached?
  • If the rate limit is reached how loang do I have to wait until I can download again?
  • Why do some Twitter user accounts work and some do not?

And so on…

My time at university was only one part about these and the rest I will probably tell in another post.

Social Media in Austria

Overview

There are a lot of statistics done for Social Media Data out there, especially for the USA users. I set out to write down a collection of these statistics for Austria. Besides this, I was also interested in finding out what can be done with Social Media Data in general.

Firsts things first, to start from the beginning, maybe it would be good to have an overview of what kind of Social Media is out there. A very good and complete overview can be downloaded at Overdrive Interactive. Not all the mentioned Social Media Platforms and Networks in the overview are used in Austria and not all of the data can be collected and used. The most recent overview statistic that I found for Austria is from CREVO Marketing & Media KG. The chart shows the number of unique users for each Social Media Platform in Austria in 2012.

Social Media Platform in Austria in 2012
Some current numbers about the most used Social Media Platforms in Austria look like these:

Social Media Platform Users Users for AT Gender specific Users Gender specific Users for AT
Twitter „Twitter users are only 18% of internet users and 14% of the overall adult population“ [1]645,750,000 Users [2] 117.431 Users68.109 active Accounts* 44.623 writing Accounts* 18.788 reading Accounts*[3] *date from 13.03.2014 „Twitter does not collect gender-based information nor return it in the API“ [4] „Natural language processing (bag of words); some words are statistically more likely to appear in a female tweet, but obviously there are no guarantees.“ [4]
Instagram ~ 150 Users million ? ? ?
Facebook 1,310,000,000 Users [2] 3.240.000 Users [3] „Facebook skews slightly toward women. But it is more gender neutral than Pinterest and Google+.“ [5] 1.580.000 Women 1.660.000 Men 2.800.000 between 14 and 49 [3]
LinkedIn 200 million Users [6] 245.000 members from Austria „Men (24%) are more likely to use LinkedIn than women (19%)“ [7]In 2011 – 76% of Users were men 24% women [8]
XING 6,1 million Users  in German speaking space [6] 2011 – 70% of Users were men, 30% women [8]
Pinterest 21 Mil. Unique Users ~ 0.8 million Users „Women are four times more likely to be Pinterest users than men.“ [9]

What is definitely missing is some statistics about Instagram in Austria.  Unfortunately I could not find anything accurate enough to post.


Social Media Data generated Topics

First of all, we can make a lot of User Statistics

  • Gender specific user statistics for different Social Media Platforms can be seen on a chart from The Statistics Portal done in 2013 for the US population.

Gender specific user statistics for different Social Media Platforms

Social Media Users by Generation in US 2010

  • Time spend on Social Media Platforms from Desktop of Mobile devices is presented in a Statista chart for US users in 2013


Social Network Activity: Mobile vs. Desktop for US in 2013

Election statistics can be done from Twitter data.

  • This is shown in a 2014 European Parliament election statistic done by twitminster.co.uk

Europe Parliament 2014 Estimations

Event prediction

Internet censorship

Youtube censorship worl history


Platform Specific Topics

From Facebook date

  • Create Social Graphs

Facebook Social Graph

On Twitter data

  • Map trending words in a specific area and of a specific timeline

Twitter trending words

Hate words heat map

  • Create a map of Tweets by Self-Identified Bankers and Artists in New York City from floatingsheep.org

Tweets by Self-Identified Bankers and Artists in New York City

Twitter posts refering to cheap beer

Mapping racism from Twitter posts

Top 15 Twitter Users Austria 2014

  • Or one could use Twitter to build an own story on Storify.com

From data posted on Flickr, Twitter and Instagram

We can answer questions like:

  • How many users of these platforms post pictures?
  • How many users of these platforms post geotagged posts and where are they?

These answers we can find on an example project from arcgis.com

Geotagged posts

Whit these being said, I will finish my examples. Feel free to leave me a comment if you know of any other such projects or any ideas you can think of for which you can use Social Media Data for.


Creative Commons License
This work by Timea Turdean is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.