1
How Stalkable Are You?
Lily R. Jenkins and Diane E. Gan
CSAFE CentreUniversity of Greenwich
2
Introduction Background to this work Overview of Tools Experiments Summary of Results Legal implications Recommendations Conclusion
C-SAFE - University of Greenwich
Contents
3
Most teenagers today have at least one “profile”
They reveal a lot of personal information about themselves that anyone can see
Their location and identity are turned on by default
Twitter users have the ‘handle’ (username) on all their social media sites
Makes it easy to identify and follow them through cyber space
C-SAFE - University of Greenwich
Introduction
C-SAFE - University of Greenwich
4
Twitter first appeared on in March 2006 Currently has 200 million active users who
send over 400 million tweets per day Added the geo-location function to user
profiles in 2009 Many users are not aware that they are
exposing their private information Enables followers to know exactly where an
individual was tweeting from The question is – do users know how to use
this feature or how to protect themselves?
Background
C-SAFE - University of Greenwich
5
Twitter’s privacy policy Clearly states that all user profiles and
subsequent tweets are by default public
Also details how the information will be used through their services such as applications, websites and third parties
Background
C-SAFE - University of Greenwich
6
Investigated a range of tools and selected:- StreamdIn, Twitonomy and Creepy
StreamdIn Application for both android and iOS Displays tweets on Google Maps using the
geo-location details attached to each tweet User’s profile picture is displayed on a map Grouped by location View numerous real-time tweets coming in
Overview of Tools
C-SAFE - University of Greenwich 7
Tracking a mobile phone
C-SAFE - University of Greenwich 8
Being Tracked on Public Transport
C-SAFE - University of Greenwich
9
Twitonomy Web based analytics tool Allow monitoring, managing and tracking
your own or another person’s activities Main feature - overall statistics of a user Includes
◦ how often they retweet◦ time of day they tweet◦ avg number of tweets sent per day◦ gives location details◦ Mentions Map - displays where in the world the
most mentions are coming from
Overview of Tools
C-SAFE - University of Greenwich 10
Twitonomy Showing Accounts From Two Different Users That Have Typical Working Days
C-SAFE - University of Greenwich
11
Creepy Aggregation program Gathers geo-location information from Twitter,
Instagram and Flickr Requires authentication with each social networking
site supported Users can be added to a target list and their geo-
location data can be retrieved ‘Current Location Details’ gives
◦ social media platform◦ time and date◦ location of the tweet◦ context of the tweet.
Using this feature it is possible to identify their current location on the map
Overview of Tools
12C-SAFE - University of Greenwich
C-SAFE - University of Greenwich
13
Subjects - three users who are prolific tweeters
Objective was to see how much information can be retrieved using freely available tools
The users will be referred to as User A, User B and User C
All have been asked to tweet with their geo-location settings turned on
Experiments
C-SAFE - University of Greenwich
14
User A and User B did not have any tweets appear on the StreamdIn map
User C’s profile picture popped up all over London
Filtered results display only one user’s tweets
Experiment 1 – StreamdIn
C-SAFE - University of Greenwich 15
Filtered view of User C’s profile picture
Shows up all over London
C-SAFE - University of Greenwich
16
Analyses the last four months’ worth of tweets
User A◦ showed information about where they tweet from◦ mostly use Twitter to re-tweet or reply◦ most activate during the winter months◦ no indication whether this user has a job
Experiment 2 – Twitonomy
C-SAFE - University of Greenwich 17
User A
Last update - 9 minutes ago
Tweet history
Platforms used
C-SAFE - University of Greenwich
18
User B◦ re-tweets and replies which suggests they use
Twitter to stay in touch with fellow users
◦ no indication as to where User B worked or lived
Experiment 2 – Twitonomy
C-SAFE - University of Greenwich 19
User B
More tweets
Significant increase in tweet history
Platforms used
C-SAFE - University of Greenwich
20
Experiment 2 – Twitonomy
User B’s Tweeting Habits
C-SAFE - University of Greenwich
21
User C revealed a distinctive pattern of usage suggests this user has a Monday to Friday job most tweets are outside of the hours of 9 to 5 it can be seen that this person has an iPhone
Experiment 2 – Twitonomy
C-SAFE - University of Greenwich
22
User A clusters of tweets can be identified single tweets showing journey information
between the clusters home address was identified by reading the
tweet content Google street easily found the house Also every Monday they attend ‘Movie Night’
at the same time and place
Experiment 3 – Creepy
23C-SAFE - University of Greenwich
User A
C-SAFE, University of Greenwich 24
The Giveaway Tweet
C-SAFE - University of Greenwich
25
User B clusters of pins identified their place of work
and their home address home residence was given away by tweets
that specifically mention the word ‘home’ Gives longitude and latitude co-ordinates
Experiment 3 – Creepy
C-SAFE - University of Greenwich 26
User B
27C-SAFE - University of Greenwich
User B’s Route to work
28C-SAFE - University of Greenwich
Locating User B’s work place
They actually only sent one tweet from work!
C-SAFE - University of Greenwich
29
User C always took the same route to work analysing the route to work showed that the second
half of the journey home may change if they needed to go to the supermarket
they never mentioned work or home in their tweets however, they were in the area of Southwark week
days between 9 and 5 only analysing each tweet and pin drop showed that they
were in Southwark every week day but never at weekends also a fixed monthly pattern - every month they
travelled to visit their parents revealed by through their tweets
Experiment 3 – Creepy
C-SAFE - University of Greenwich
30
User C visit’s his parent’s house in Southampton once per month
31C-SAFE - University of Greenwich
User C’s Tweets, which establish a pattern of clusters around home and work
32C-SAFE - University of Greenwich
Three times per week User C goes to this gymWeek days between 7 and 10Weekends between 1 and 3
C-SAFE - University of Greenwich
33
How much did each users’ Tweeting expose the rest of their social media “presence”?
Did the three users have accounts on Facebook, LinkedIn, Foursquare and Instagram?
User A gave no indication that they had any other social media accounts
A Google search revealed their Facebook page The profile pictures confirmed this Logging into a Facebook account that is not
“friends” with User A gave a small number of their pictures, as well as where they were living
Experiment 4 – Other Social Networks
C-SAFE - University of Greenwich
34
User A also had a profile on Instagram
using Instagram24.com and User A’s profile name it was possible to locate their pictures
including some pictures that they had “liked”
Also found them on LinkedIn
Google Street View located their front door
Experiment 4 – Other Social Networks
C-SAFE - University of Greenwich
35
A Google search for User B found their Linkedin, Facebook and Google+ accounts
Using these profiles, it was possible to confirm ◦where they worked◦the city they live in◦where they were studying
Experiment 4 – Other Social Networks
C-SAFE - University of Greenwich
36
User C was the easiest to identify with Twitter
But the most difficult to locate on other social media sites
Only Foursquare revealed their location Back to Twitter After conducting an exhaustive search of
their Twitter account two tweets were found with pictures
Experiment 4 – Other Social Networks
C-SAFE - University of Greenwich
37
Tweet 1 Posted while in
hospital Hospital ID tag
revealed their surname their date of
birth NHS ID
Experiment 4 – Other Social Networks
C-SAFE - University of Greenwich
38
Tweet 2 e-ticket showed
their full name (including a middle name)
airports they will pass through
how long they will be stopping at each location
A gift to a burglar
Experiment 4 – Other Social Networks
C-SAFE - University of Greenwich
39
Data Protection Act (1998) states that the “data subject has given his
consent to the processing” of personal data
does not offer any conclusive reasoning as to how social networking sites users are protected
by signing up to these sites and using them in a public manner the user has given their consent
Legal Implications
C-SAFE - University of Greenwich
40
Employers may check your personal life using social networks
Example - Kent Police Commissioner’s Youth Advisor Paris Brown
forced to withdraw when her twitter content was made public
Ref: http://www.dailymail.co.uk/news/article-2312044/Paris-Brown-Foul-mouthed-youth-commissioner-quit-offensive-tweets-questioned-police-caution.html
Implications
C-SAFE - University of Greenwich
41
Reduce your risk◦ Do not tweet where you live, even if it is only the
city◦ Do not provide your phone number◦ Avoid using full names ◦ Avoid using a profile picture◦ Set your profile to private ‘Protect my Tweets’◦ Remove geo-location tagging on tweets◦ Remove “Let others find me by my email address”◦ Do not connect your Twitter account to any other
social media sites◦ Limit the amount of apps that have access to your
profile◦ Be very selective about what you put in your tweets
Recommendations
C-SAFE - University of Greenwich
42
There are a huge number of tools that retrieve your information
All tools are freely available StreamdIn, Twitonomy and Creepy were used
for these experiments Creepy was the most successful It was the geo-location data AND the tweet
contents that leaked information
Conclusion
C-SAFE - University of Greenwich
43
Questions?
Lily Jenkins [email protected]
Diane Gan [email protected]