TY - GEN
T1 - Use of supervised learning to predict directionality of links in a network
AU - Soundarajan, Sucheta
AU - Hopcroft, John E.
PY - 2012
Y1 - 2012
N2 - Often, the information contained in network data is incomplete. Many avenues of research are aimed at addressing this incompleteness. For example, the link prediction problem attempts to identify which missing links are most likely to exist in the complete network. In this paper, we consider a related, but different, problem: predicting the directions of links in a directed network. We treat this problem as a supervised learning problem in which the directions of some edges are known. We calculate various features of each known edge based on its position in the network, and use a Support Vector Machine to predict the unknown directions of edges. We consider four networks, and show that in each case, this method performs significantly better than other compared methods.
AB - Often, the information contained in network data is incomplete. Many avenues of research are aimed at addressing this incompleteness. For example, the link prediction problem attempts to identify which missing links are most likely to exist in the complete network. In this paper, we consider a related, but different, problem: predicting the directions of links in a directed network. We treat this problem as a supervised learning problem in which the directions of some edges are known. We calculate various features of each known edge based on its position in the network, and use a Support Vector Machine to predict the unknown directions of edges. We consider four networks, and show that in each case, this method performs significantly better than other compared methods.
UR - http://www.scopus.com/inward/record.url?scp=84872692487&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84872692487&partnerID=8YFLogxK
U2 - 10.1007/978-3-642-35527-1_33
DO - 10.1007/978-3-642-35527-1_33
M3 - Conference contribution
AN - SCOPUS:84872692487
SN - 9783642355264
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 395
EP - 406
BT - Advanced Data Mining and Applications - 8th International Conference, ADMA 2012, Proceedings
T2 - 8th International Conference on Advanced Data Mining and Applications, ADMA 2012
Y2 - 15 December 2012 through 18 December 2012
ER -