Ranking Answering in Social Networks

Qing Yang,Ping Zhou,Huibing Zhang,Jingwei Zhang
DOI: https://doi.org/10.1109/iciss.2010.5655510
2010-01-01
Abstract:Users in social networks can follow and be followed to establish all kinds of cliques. Users' rankings in followed numbers show their popularities, which is an important measurement for social network analysis. In order to give a fast answering for ranking queries, we proposed four methods, ISH, HDD, BBT and SMH. ISH uses inverted sequential hash, HDD is a combination of blocking and hierarchy, BBT utilizes blocking and B- tree technologies, SMH analyzes the sparse of data in large number space, and adapts sequential storage for low values and multi-hashing storage for high values. Our experiments present their different performance in query time and memory consuming.
What problem does this paper attempt to address?