Aug 12, 2007WebKDD and 1st SNA-KDD Workshop Why We Twitter? Understanding Microblogging Usage and Communities Akshay Java Tim Finin University of Maryland,

Aug 12, 2007 WebKDD and 1st SNA-KDD Workshop

Why We Twitter?Understanding Microblogging Usage and

CommunitiesAkshay Java

Tim Finin

University of Maryland, Baltimore County

Xiaodan Song

Belle Tseng

NEC Laboratories America Inc.

Aug 12, 2007 Akshay Java

What is Twitter?

Micro-Blogging, Social Networking Service Users send updates/tweets via SMS, IM or Web Light-weight blogging, short posts

(140 characters or less)


What is Twitter?

Easily share status messagesEasily share status messages

Twitter post

Current Status

Friends


MICRO-BLOGS


Ambient Intimacy

“Ambient intimacy is about being able to keep in touch with people with a level of regularity and intimacy that you wouldn’t usually have access to, because time and space conspire to make it impossible. “

- Leisa Reichelt Twitter, Flickr and other Social Media sites.


Twitter! Twitter! Twitter!

Disclaimer: No association to Twitter Inc/ Obvious corp.Disclaimer: No association to Twitter Inc/ Obvious corp.


Motivation

Goal: Study Micro-blogging usage, user-intentions and community structure.

Motivation:– What is the excitement about?– What is the user-intention in micro-blogging?– How is this form of communication different?


Outline

Why We Twitter?Why We Twitter?

Micro-blogging Usage


User IntentionsUser Intentions

Community-based Intentions


Content-basedIntentions



Social Media

Social media describes the online technologies and practices that people use to share opinions, insights, experiences, perspectives with each other.

~ Wikipedia 07


Social Media

Social media describes the online technologies and practices that people use to share opinions, insights, experiences, perspectives and engage with each other.

Examples:

Blogs, Wikis, Flickr, YouTube, Micro-blogs…


Outline









USAGE

• Popularity of Twitter?

• Is it very different from blogs?

• What is it’s adoption across the world?


Dataset Description

Constructed by monitoring the Public timeline of posts 2 month period (04/07-05/07) 1,348,543 posts 76,177 Users 829,053 friend relation between them Used the Twitter API for accessing user social network


The Twitter Phenomenon

Twitter’s popularity increased after winning the Web award at SXSW conference, March 2007.

Currently, Twitter is the most popular micro-blogging tool.Currently, Twitter is the most popular micro-blogging tool.Source: ComScore


Growth Rate

Slow down in growth of new users joining the network.

Slow down in growth of new users joining the network.

Dedicated user-base generating new updates/tweets.

Dedicated user-base generating new updates/tweets.


Network Statistics

Degree distributions similar to Blogs, Web…Degree distributions similar to Blogs, Web…


Network Statistics

However, higher reciprocal linking and clustering coefficient indicates mutual acquaintance.However, higher reciprocal linking and clustering coefficient indicates mutual acquaintance.

Network Statistics


User Retention

ACTIVE

RETAINED

User has at least one post in the week

Active user, who reposts the following week

Shows a continued activity and retention on TwitterShows a continued activity and retention on Twitter


Geographical SpreadCross Continent Social Network

Continental Network Properties

Top Cities: Tokyo, NY, SF, Seattle, LA, Chicago, Toronto, Austin, Singapore, Madrid

Global popularityGlobal popularity

Social network crosses geographical boundaries

Social network crosses geographical boundaries

Higher reciprocity in Europe and AsiaHigher reciprocity in Europe and Asia


Outline









USER INTENTION

• What are we using Twitter for?

• Are there communities?

• If so, what are the community-level intentions?


User IntentionsUsing Link Structure:

– Information source Such users have a number of followers ( include bots like forecast, stock, CNN breaking news, etc.)

– Information seeker Such users may post infrequently, but have a number of connections

– Friendship relation Most user’s social network is within mutual acquaintances

Using Content:– Daily chatter dinner, work, movie…

– Conversations (@) Reply to a specific person @ev

– Sharing URLs Sharing URLs through tinyURL

– Commenting on News Number of automated RSS to Twitter bots posting news


Communities in Twitter First find Hubs and

Authorities using HITS Consider only bidirectional

links Clique Percolation Method

(CPM) to find overlapping communities

A-list bloggers and personalities are on Twitter

Scobleizer, SteveRubel, JasonCalacanis, SteveJobs

JohnEdwards, BarakObama, et al.

Hubs and Authorities


Clique Percolation Method (CPM)

Basic Idea

Two nodes belong to the same community if they can be connected through adjacent k-cliques.

I. Derenyi, G. Palla, and T. Vicsek. Clique percolation in random networks. Physical Review Letters, 94:160202, 2005.G. Palla, I. Derenyi, I. Farkas, and T. Vicsek. Uncovering the overlapping community structure of complex networks in

nature and society. Nature, 435:814, 2005.

Example Gaming Community


com:175 twitter:134just:133 like:86good:82 tinyurl:75time:74 new:74jasona:73 going:68day:63 don:61work:58 think:56ll:54 scottw:54today:52 hkarthik:50nice:49 getting:47got:47 really:46yeah:44 need:43watching:41 love:41night:40 home:40

com:93 twitter:74just:35 new:32tinyurl:29 going:24ll:22 blog:21jaiku:21 don:21leo:21 flickr:21like:19 video:18google:18 today:18feeds:18 getting:16yeah:16 good:15people:15

com:93 twitter:76 tinyurl:34just:32 new:28 video:26going:24 ll:22 jaiku:22blog:21 leo:21 like:19don:19 gamerandy:19 yeah:18google:17 live:16 people:16got:16 know:15 time:15

com:121 twitter:76 just:50ustream:43 tv:42 live:42today:39 hawaii:36 day:33new:33 time:33 good:33video:32 leo:30 work:30like:28 watching:28 tinyurl:28

com:198 twitter:132 just:109tinyurl:87 going:59 blog:56like:55 good:51 new:50url:50 day:49 people:46time:45 today:45 google:42don:41 think:40 night:38ll:38 need:35 got:33ireland:33 great:31 looking:29work:29 thanks:28 video:26

INFORMATION

HUB

Information Source: Communities connected via Robert Scoble, an A-list bloggerInformation Source: Communities connected via Robert Scoble, an A-list blogger


Key Termsgoing:222 just:218work:170 night:143bed:140 time:139good:137 com:130lost:124 day:122home:112

listening:111 today:100 new:98got:97 gspn:92watching:92 kids:88morning:81 twitter:79getting:77 tinyurl:75lunch:74 like:72podcast:72 watch:71ready:70 tv:69need:64 live:61tonight:61 trying:58love:58 cliff:58dinner:56

Key Terms just:312 com:180

work:180 time:149listening:147 home:145going:139 day:134got:126 today:124good:116 bed:114night:112 tinyurl:97getting:88 podcast:87dinner:85 watching:83like:78 mass:78lunch:72 new:72ll:70 tomorrow:69ready:64 twitter:62working:61 tonight:61morning:58 need:58great:58 finished:55tv:54

INFORMATION

BRIDGE

Information Source, Information Seeker: Different roles in different communitiesInformation Source, Information Seeker: Different roles in different communities


STAR NETWORKS /

SMALL CLIQUES

Friendship-relation: Small groups among friends/co-workersFriendship-relation: Small groups among friends/co-workers


Outline









USER INTENTION

• What are the distinctive terms?

• Are there any trends?


Wisdom of the Crowds

log-likelihood ratio

Popular topics: Activities, Current Events, TV shows/EntertainmentPopular topics: Activities, Current Events, TV shows/Entertainment


Twittermenthttp://twitterment.umbc.eduSearch and Trend analytics on TwittersphereSearch and Trend analytics on Twittersphere

lunch dinnerwork

lunch

dinner

coffee

chipotle

panera

http://twitterment.umbc.edu/


Outlook on Micro-Blogging

The future is here!

Twitter,Jaiku,Pownce,FaceBook Status message is now public

– Ambient intimacy– Information sharing

Thoughts on new applications– Users can play different roles in different communities

– Number of updates received can be quite overwhelming

– New tools and services would benefit from allowing greater personalization based on user-intentions

e.g. Separating work and friend social network


Conclusions First study of the Microblogging phenomenon. Popularity of Micro-blogging due to the combined benefits of

– Light-weight blogging – and the ability to share information in the social network.

Main user-intentions– Information sharing– Information seeking– Friendship– Users generated content includes:

Status updates, daily chatter, sharing links/News, etc. Future Work

– Automatically classifying the different intentions.– Finding frequent patterns in community structure.


Thank You! Questions?

Thanks to Twitter Inc. for providing the Twitter API.

http://twitter.com/akshayjava


Backup Slides


User Intentions

Aug 12, 2007WebKDD and 1st SNA-KDD Workshop Why We Twitter? Understanding Microblogging Usage and Communities Akshay Java Tim Finin University of Maryland,

Documents

akshay java twitter

akshay java microblogs

akshay java user rete

web slide

akshay java motivation

user social network

comscore slide

network statistics slide