An update algorithm for restricted random walk clustering for dynamic data sets

Abstract

Clustering, Dynamic clustering, Random walks, 60J20, 62H30,

Community detection in Social Media

“…a balance is achieved between temporal noise and concept drift (Chakrabarti et al 2006). Incremental methods can be further divided in iterative approximation schemes (Sun et al 2007;Yang and Liu 2006;Lin et al 2008) and graph modification oriented reclusterings (Falkowski et al 2007;Franke and Geyer-Schulz 2009).…”

“…For example, the DenGraph framework (Falkowski et al 2007), which uses a notion of subgraph density equivalent to the concept of (μ, )-cores (Xu et al 2007), translates each modification taking place on the graph into some update operation (or no operation in some cases) on the graph cores and their associated communities. A more sophisticated scheme is presented by Franke and Geyer-Schulz (2009) based on an incremental version of a restricted random walk clustering scheme, which takes into account only the changes that have taken place on the graph since the last clustering.…”

Emerging, Collective Intelligence for Personal, Organisational and Social Use

“…Processing a network with about 5 million nodes and 40 million edges takes less than 10 seconds on standard desktop computers. Another algorithm we developed is optimized for the processing of huge dynamic datasets [72]. The algorithm is based on restricted random walks that are incrementally updated.…”

Μέθοδοι Εξόρυξης Γνώσης Σε Συλλογές Περιεχομένου Μεγάλης Κλίμακας Από Κοινωνικά Δίκτυα

“…συν., 2007) µεταϕράζει κάθε µεταβολή στο γράϕο σε µια πράξη ανανέωσης (update operation) στους (µ, ϸ)-πυρήνες του γράϕου και τις κοινότητες που σχηµατίζονται γύρω από αυτούς. Στην εργασία (Franke & Geyer-Schulz, 2009) υλοποιείται µια επαυξητική εκδοχή του αλγορίθµου πεπερασµένων τυχαίων περιπάτων (Restricted Random Walks, RRW) που λαµβάνει υπόψη µόνο τις µεταβολές που έλαβαν χώρα µεταξύ του παρόντος και του προηγούµενου στιγµιότυπου γράϕου. (Tyler, Wilkinson, & Huberman, 2003) στα πλαίσια της µεθόδου GN (Girvan & Newman, 2002).…”