Home | Syllabus | Course Material | Assignments | Golbeck Home | INFM 289I: Social Networks |
Karrie Karahalios on Tie Strength and Text
It may also be useful to skim this paper, by the same person, on computing tie strength from Facebook profiles.
Eric Gilbert and Karrie Karahalios. 2009. Predicting Tie Strength With Social Media. Proceedings of CHI 2009.
In all of this work, Dr. Karahalios talks about how tie strength can be inferred from data, especially text. I want you to do a brief exercise. Spend around 15 minutes on it. Imagine I have sent you to a public place - a park, the mall, etc. Without listening in on conversations, I want you to observe things about people to guess their tie strength. In class, we talked about general observations - body language, physical contact, etc. For this exercise, I want you to come up with things to measure. For example, we may watch two people having a conversation and time how long each person talks. That could tell us if one person dominates the conversation, which may be informative. We may count the number of times people touch one another.
For each of the seven factors relating to tie strength - Intensity, Intimacy, Time, Reciprocal Services, Structural, Emotional Support and Social Distance - list three things you could measure from observing people interact in public.
Write an email (please do not include an attachment - put everything directly in the email) with your exercise and email it to be at jgolbeck@umd.edu. Use the EXACT SUBJECT LINE "289I Exercise 2". Please turn this in by 4pm Friday 3/2.