Stanford University professor Jure Leskovec is studying the traces human activity leaves on the Internet and how that data can be used to analyze human behavior on the Web.
Based on information about the personal networks of Facebook users and their communications, Leskovec was able to predict half of the new contacts users would add in the near future. Eventually, the rate of correctly predicted new friends could be even higher, Leskovec says.
The data shows that a social network should neither be too sparsely populated nor too dense or else it could hinder the network's growth. The research could be used to create models of how online groups develop over time.
Leskovec also is studying Microsoft Instant Messenger to prove the six degrees of separation hypothesis. He found that people using the program were about 6.6 steps apart from each other, which could be useful for determining the shortest path between two computers on the Internet.
Leskovec also wants to develop algorithms that find patterns for how new media is integrated into the Web, which could show how information changes gradually over time.
From IDG News Service
View Full Article
No entries found