| Sun | Mon | Tue | Wed | Thu | Fri | Sat |
|---|---|---|---|---|---|---|
| 1 | 2 | |||||
| 3 | 4 |
5 |
6 |
7 | 8 |
9 |
| 10 | 11 | 12 | 13 | 14 | 15 | 16 |
| 17 | 18 | 19 |
20 | 21 | 22 | 23 |
24 |
25 |
26 |
27 | 28 |
29 |
Editor Login
Convener in chief:
David Lazer
(Methodology, Networked Governance)
Editors:
Stanley Wasserman
(Current Trends, Methodology, Social Networks)
Allan Friedman
(Simulations)
Nathan Eagle
(Technology, Social Computing, Powerlaws, Current Trends)
Ben Waber
(Technology, Social Computing) Thomas Langenberg
(Technology, Social Computing, Social Networks, Current Trends)
Ines Mergel
(Knowledge Sharing, Social Computing, Social Software, Current Trends)
Brian Rubineau
(Social Dynamics, Societal Networks, Simulations)
Maria Binz-Scharf
(Qualitative Methodology, Knowledge Sharing, eGovernment)
Jeff Boase
(Technology, Societal networks)
Alexander Schellong
(Admin, eGovernment, Citizen Relationship Management)
Categories
Business
Citizen Relationship Management
Commentary
Complexity
Current Trends
Events/Announcements
Innovation
Internet
Knowledge Sharing
Methodology
Networked Governance
Networks in political science
PNG Working Paper Series
Powerlaws
Reviews
SN in the news
Simulations
Small Worlds
Small group networks
Social Capital
Social Computing
Social Finance
Social Networking Platforms
Social Psychology
Social Software
Societal Networks
Technology
eGovernment
Archives
- February 2008
- January 2008
- December 2007
- November 2007
- October 2007
- September 2007
- August 2007
- July 2007
- June 2007
- May 2007
- April 2007
- March 2007
- February 2007
- January 2007
- December 2006
- November 2006
- October 2006
- September 2006
- August 2006
- July 2006
- June 2006
- May 2006
- April 2006
- March 2006
- February 2006
- January 2006
- December 2005
- November 2005
Recent Entries
- Interview: Thorsten Jacobi on the current state and trends in social software
- Online Social Network User Fatigue?
- Second Call for Papers / Funding for doctoral student training for Harvard conference on networks
- Obama, techology, and social networks
- Tapping on the wisdom of the crowd: Social network analysis software tools on Wikipedia
Recent Comments
- Re: Automatically Sensing Social Networks
By: Ben Waber
Date: May 3, 2007 11:29 AM - Re: Automatically Sensing Social Networks
By: David Allen
Date: May 3, 2007 11:12 AM
Notification
« eGooglement - How google is trying to improve the accesibility of government websites | Main | Wayne Baker on Social Capital and Generalized Reciprocity »
3 May 2007
Automatically Sensing Social Networks
My group at the MIT Media Lab has started conducting experiments using the Sociometric Badge, a sensing platform that logs voice features, proximity to other individuals, face-to-face interactions, and movement. We have finished analysis of data obtained in a preliminary study at a German bank’s marketing division, and the results were astonishing. We were also fortunate enough to get the e-mail logs for the bank over the course of the study.
For years e-mail data has been used as an easy to obtain proxy for social network information. We found, however, that in fact proximity is highly negatively correlated with e-mail use. We can attribute this to several factors. First, if you are in close proximity to another individual, it makes more sense to interact with them in the real world rather than send them an e-mail. Second, proximity information also picks up on informal relations, while in this particular organization e-mail is used mainly for business purposes. This is because if you spend a lot of time with someone you are more likely to be their friend and therefore less likely to send an e-mail to them. This result points towards the necessity of having face-to-face interaction information in order to have a full view of the social network.
When we combined face-to-face interaction information (which we used to create a social network representation) with e-mail data we got a more complete view of the social network that exhibited some interesting properties. Most notably, betweenness in the social network and total communication were both highly negatively correlated with the perceived quality of interactions. These two measures, however, were not significantly correlated with each other. This points to the necessity of aiding central actors in managing their interaction-related stress, since it is evident that those who are overburdened with their communication responsibilities feel that their interaction quality similarly degrades.
These results strongly argue for the use of automatic sensing data for computational social science. We report many other interesting findings in a paper that will appear in the NetSci 2007 conference.
Posted by Ben Waber at May 3, 2007 10:18 AM
-
.
- Business (2) .
- Citizen Relationship Management (13) .
- Commentary (5) .
- Complexity (2) .
- Current Trends (40) .
- eGovernment (15) .
- Events/Announcements (51) .
- Innovation (12) .
- Internet (14) .
- Knowledge Sharing (16) .
- Methodology (28) .
- Networked Governance (14) .
- Networks in political science (1) .
- PNG Working Paper Series (6) .
- Powerlaws (4) .
- Reviews (2) .
- Simulations (1) .
- Small group networks (3) .
- Small Worlds (2) .
- SN in the news (56) .
- Social Capital (2) .
- Social Computing (9) .
- Social Finance (1) .
- Social Networking Platforms (8) .
- Social Psychology (2) .
- Social Software (22) .
- Societal Networks (33) .
- Technology (27)
Comments
Ben, can you say what were some of the more astonishing results, if anyway those go beyond your last full para in the post? Thanks.
Posted by: David Allen at May 3, 2007 11:12 AM
I don't want to post the entire paper here, but in particular I believe that the results regarding e-mail are quite important since SN research has relied so heavily on this information in the past. Another very interesting result was that communication between managers and employees was very highly negatively correlated with perceived interaction quality, implying that having more interactions with either your subordinates or your boss is draining.
Surprisingly, e-mail and proximity ties were highly correlated when communication was present and both individuals were of the same hierarchical level (employee-employee and manager-manager). This correlation implies that e-mail communication between individuals with the same role increases as their proximate time increases. Causality in the other direction is also possible, with high e-mail communication breeding an increase in proximity through meetings and the like.
These are probably some of the biggest results, but there are others that will appear in the paper. There's just too many to list in a post! We have also submitted a paper to ISWC that contains results using movement and speech data.
Posted by: Ben Waber at May 3, 2007 11:29 AM