Measuring Serendipity: Connecting People, Locations and Interests in a Mobile 3G Network Ionut Trestian Supranamaya Ranjan Aleksandar Kuzmanovic Antonio.

Measuring Serendipity: Connecting People, Locations

and Interests in a Mobile 3G Network

Ionut TrestianSupranamaya RanjanAleksandar KuzmanovicAntonio Nucci

Northwestern UniversityNarus Inc.

http://networks.cs.northwestern.edu http://www.narus.com

Ionut TrestianMeasuring Serendipity: Connecting People, Locations and

Interests in a Mobile 3G Network

2

Social network websites among the most popular websites on the Internet

Online Social Networks

Ionut TrestianMeasuring Serendipity: Connecting People, Locations and

Interests in a Mobile 3G Network

Mobage Town

Japan based mobile social network

11 million users

Allows users to:– Send messages, chat in

communities, exchange music, read pocket novels, write blogs, play games etc.

3

Ionut TrestianMeasuring Serendipity: Connecting People, Locations and

Interests in a Mobile 3G Network

Loopt

4

Allows contacts to visualize one another’s location using mobile phones and share information

Available for Sprint, Verizon, At&t, T-Mobile on devices such as BlackBerry, iPhone and gPhone

Ionut TrestianMeasuring Serendipity: Connecting People, Locations and

Interests in a Mobile 3G Network

Other Location Based Services

Sharing your location with friends (BuddyBeacon –for iPhone)

Location based searches (EarthComber)

Notifications about places and events around you (LightPole)

Tagging locations (Metosphere)

5

Ionut TrestianMeasuring Serendipity: Connecting People, Locations and

Interests in a Mobile 3G Network

Research Questions

6

How likely are we to meet in our daily lives people who share common interests in the cyber domain?

What is the relationship between mobility properties, location, and application affiliation in the cyber domain?

3,162,818 packet data sessions generated by 281,394 clients in 1196 locations (Base Stations) across a large

metropolitan area

Ionut TrestianMeasuring Serendipity: Connecting People, Locations and

Interests in a Mobile 3G Network

Extracting Human Movement

7

1. Intra-session movement

RADA Start(contains BSID)

RADA Update(contains BSID)

2. Inter-session movement

RADA Stop(contains BSID)

RADIUSServer

BaseStation 1

BaseStation 2

Note that we have only a sampled view of human movement.

How well can we do?

Ionut TrestianMeasuring Serendipity: Connecting People, Locations and

Interests in a Mobile 3G Network

Extracting Human Movement

8

Despite sampled observations we still do a good job at understanding user movement.

The ordering of the curves accounts for the larger time span which can accommodate larger travel distances

Most human movement is over short distances.

Ionut TrestianMeasuring Serendipity: Connecting People, Locations and

Interests in a Mobile 3G Network

Extracting Application Interest

9

http://www.singlesnet.com

Dating website

http://www.facebook.com

Social networkingwebsite

http://www.mp3.com

Music downloadwebsite

Interest Keywords

Dating dating, harmony, personals, single, match

Music song, mp3, audio, music, track, pandora

Social netw. facebook, myspace, blog

Keyword based URL mining

Ionut TrestianMeasuring Serendipity: Connecting People, Locations and

Interests in a Mobile 3G Network

Rule Definitions

10

Ionut TrestianMeasuring Serendipity: Connecting People, Locations and

Interests in a Mobile 3G Network

Rule Mining

11

Location A

Location B(A, B, w, δ)

Rule support:Number of people

present at A

Rule confidence:Number of people that

move from A to B

Rule confidence probability:confidence/support

Wδ

Ionut TrestianMeasuring Serendipity: Connecting People, Locations and

Interests in a Mobile 3G Network

12

Rule Statistics

Total co

nfid

ence o

f rules

Increase in number of active users at commute hours (8AM and 5PM)

Movement rules are more active during day time, also less active during weekend

Ionut TrestianMeasuring Serendipity: Connecting People, Locations and

Interests in a Mobile 3G Network

Location Rank – Application Accesses

13

Music downloads – anti-correlation with mobility spanMail – correlation with mobility spanSocial netw. – dominates the medium mobility range

Ionut TrestianMeasuring Serendipity: Connecting People, Locations and

Interests in a Mobile 3G Network

Location Ranking

14

All users spend most of their time in their top 3 locations

Comfort zone

3

Ionut TrestianMeasuring Serendipity: Connecting People, Locations and

Interests in a Mobile 3G Network

Location Rank – Application Accesses

15

Music downloads, Dating, Trading heavily accessed in the comfort zone

Comfort zone

Social netw. News and Mail tend to be accessed outside too

Note that Dating is accessed more in the Comfort Zone

Ionut TrestianMeasuring Serendipity: Connecting People, Locations and

Interests in a Mobile 3G Network

Home vs. Work

Ionut TrestianMeasuring Serendipity: Connecting People, Locations and

Interests in a Mobile 3G Network

Hotspots

Via rule mining we detect highly active locations

17

We identify 4 types of such locations

– Noon hotspots – 28 such locations• Highly active during Noon hours

– Night hotspots – 62 such locations• Highly active during night hours

– Day-office hotspots – 23 such locations• Highly active during day hours

– Evening hotspots – 8 such locations• Highly active during the evening

Ionut TrestianMeasuring Serendipity: Connecting People, Locations and

Interests in a Mobile 3G Network

Biased Application Access at Hotspots

18

Applicationaccesseshotspots

Normalized user affiliation

Despite similar userbase at hotspots during the seven day interval, application accesses are highly skewed

towards certain applications.

Ionut TrestianMeasuring Serendipity: Connecting People, Locations and

Interests in a Mobile 3G Network

Application Access - Time of Day

19

Applicationaccessesnon hotspot times

Applicationaccessesnon hotspots

However the bias in application access is not entirely due to an illusive “time of day” effect !

Ionut TrestianMeasuring Serendipity: Connecting People, Locations and

Interests in a Mobile 3G Network

20

Regional analysis – Spectral Clustering

Using spectral clustering we:

Cluster locations as belonging to regions

Cluster users as belonging to regions

Spectral clustering doesn’t make any assumptions on the shape of the clusters(opposed to k-means)

Ionut TrestianMeasuring Serendipity: Connecting People, Locations and

Interests in a Mobile 3G Network

Clustering Results

Ionut TrestianMeasuring Serendipity: Connecting People, Locations and

Interests in a Mobile 3G Network

Regional Analysis – Research issues

Two relevant issues for location based services:

– Time independent interactions(useful for tagging services) – part of user trajectories overlap irrespective of the time of the movement

– Time dependent interactions – same location same time

Questions:– How many distinct people with the same interests do we

meet?• Strongly dependent on userbase (probability to meet people

higher in clusters with bigger userbase)

– How often do we meet people?

22

Ionut TrestianMeasuring Serendipity: Connecting People, Locations and

Interests in a Mobile 3G Network

Time Independent Interactions

23

Cluster 1 has a higher number of interactions per location mainly because of larger hotspot density

27/162 (Cluster 1)> 26/257 (Cluster 4) for night hotspots

Ionut TrestianMeasuring Serendipity: Connecting People, Locations and

Interests in a Mobile 3G Network

24

Who Will Win the Interaction Race?Event type Mobile users

Seen in more than 20

locations

Static users(hotspot)

Spent more than 6 hours in

a Hotspot

Static users(non-hotspot)Spent more

than 6 hours in a non-hotspot

Social netw. 704 604 424

Music 828 565 319

Dating 253 188 96

Mobile users clearly win the interaction raceHowever it pays off to spend time in popular locations

Ionut TrestianMeasuring Serendipity: Connecting People, Locations and

Interests in a Mobile 3G Network

Conclusions

First study at such a large scale aimed at correlating mobility, location, and application usage

Provided new insights from user perspective, location perspective, and provider perspective that shows the enormous location based service potential

25