TY - GEN
T1 - Avatar path clustering in networked virtual environments
AU - Jiang, Jehn Ruey
AU - Huang, Ching Chuan
AU - Tsai, Chung Hsien
PY - 2010
Y1 - 2010
N2 - With the increase of network bandwidth and the advance of 3D graphics technology, networked virtual environments (NVEs) have become popular recently. Early SIMNET and currently booming massively multiplayer online games (MMOGs), such as Second Life (SE) and World of War craft (WoW), are examples of NVEs. Because NVE users' interests or habits may be similar, avatars, or the representative of NVE users, may have similar behavior patterns, which leads to similar motion paths in the NVE. This paper proposes two NVE avatar path clustering algorithms, namely, Average Distance of Corresponding Points-Density Clustering (ADCP-DC) and Longest Common Subsequence-Density Clustering (LCSS-DC). Given avatar paths, both algorithms will produce a collection of path clusters and their representative paths (RPs), which can be used to analyze avatar behaviors for improving NVE design. We take SE user trace data as input of the algorithms to demonstrate their applicability. We also show how to adjust algorithm parameters to obtain high-quality path clustering in terms of silhouette coefficient and cluster coverage.
AB - With the increase of network bandwidth and the advance of 3D graphics technology, networked virtual environments (NVEs) have become popular recently. Early SIMNET and currently booming massively multiplayer online games (MMOGs), such as Second Life (SE) and World of War craft (WoW), are examples of NVEs. Because NVE users' interests or habits may be similar, avatars, or the representative of NVE users, may have similar behavior patterns, which leads to similar motion paths in the NVE. This paper proposes two NVE avatar path clustering algorithms, namely, Average Distance of Corresponding Points-Density Clustering (ADCP-DC) and Longest Common Subsequence-Density Clustering (LCSS-DC). Given avatar paths, both algorithms will produce a collection of path clusters and their representative paths (RPs), which can be used to analyze avatar behaviors for improving NVE design. We take SE user trace data as input of the algorithms to demonstrate their applicability. We also show how to adjust algorithm parameters to obtain high-quality path clustering in terms of silhouette coefficient and cluster coverage.
KW - Density clustering
KW - Massively multiplyer online games
KW - Networked virtual environments
KW - Path clustering
UR - http://www.scopus.com/inward/record.url?scp=79951728892&partnerID=8YFLogxK
U2 - 10.1109/ICPADS.2010.124
DO - 10.1109/ICPADS.2010.124
M3 - 會議論文篇章
AN - SCOPUS:79951728892
SN - 9780769543079
T3 - Proceedings of the International Conference on Parallel and Distributed Systems - ICPADS
SP - 845
EP - 850
BT - Proceedings - 16th International Conference on Parallel and Distributed Systems, ICPADS 2010
T2 - 16th IEEE International Conference on Parallel and Distributed Systems, ICPADS 2010
Y2 - 8 December 2010 through 10 December 2010
ER -