Much of human interaction these days is electronically mediated, allowing the study of social networks on an unprecedented scale and with great detail. Information that could previously only be obtained through time consuming mail, phone, and face-to-face surveys, is now readily available for harvesting as a side-effect of individuals leading digital lives. We found, for example, that networks of personal homepages can give useful insights into real-world social interaction. Studies of homepage networks at two universities revealed patterns of social contact and showed to what extent shared characteristics can be used to predict whether two people know one another. We also observed that the similarity of homepages decreases as the number of hyperlinks separating them increases.
We further applied our methodology to a study of an online community consisting of over 2,000 users at a single university. The profiles, being more structured than the free-form text of homepages, allowed us to identify the importance of specific attributes such as year in school, major, personality, or favorite music genres to friendship formation. We were able to correlate chosen majors to personalities and free-time activities. Even more interestingly, we could correlate users' perceptions of themselves with how others perceive them, through a rating system designed to entertain the community by calculating who has the most social 'karma'.
We were further able to apply the insights gleaned from studying social networks to model the flow of information in social groups. Unlike viruses, which spread indiscriminately from host to host, pieces of information are propagated by people who find them interesting and who pass the information to others who they think may be interested. Since people are most similar to their immediate contacts, and this similarity decays as the distance in the social network between individuals increases, information becomes less relevant further away from the source and is unlikely to spread throughout the network. This holds true even in networks with power-law connectivity distributions where highly connected individuals, known as hubs, have the opportunity to potentially spread information to a large number of people. Taking advantage of the fact that significant portion of communication, especially within a corporation, occurs via email, we confirmed our model by observing the occurrence of attachments in mailboxes and by simulating the spread of information using actual email patterns.
Contact:
Lada Adamic
email: lada.adamic@hp.com
For more information about this and related work, please see the Information Dynamics Web site http://www.hpl.hp.com/research/idl/.
|
printable version
