***** To join INSNA, visit http://www.insna.org ***** Hello As I understand, the p1, p2, p* models are used to model network ties as independent variable, possibly using actor attributes as explanatory variables. Could you please point me to work/method that addresses the reverse problem of modeling attribute data from the network structure? In this case the attributes are dependent variables and should be explained by the network ties plus attribute data of alters and other actors in the network. I am especially interested in predicting categorical data with large number of categories. My background is in computer science and I am looking at using social networks for collaborative filtering. I would be interested in how social network scientists have approached or would approach this problem. I'll post a summary of the answers to the list - please let me know if you want me to keep your reply private. Thank you for your advice, Vaclav _____________________________________________________________________ SOCNET is a service of INSNA, the professional association for social network researchers (http://www.insna.org). To unsubscribe, send an email message to [log in to unmask] containing the line UNSUBSCRIBE SOCNET in the body of the message.