PASCAL - Pattern Analysis, Statistical Modelling and Computational Learning

An implementation of the Pathfinder algorithm for sparse networks and its application on text networks
Vladimir Batagelj, Vid Podpečan and Anže Vavpetič
In: 12. mednarodna multikonferenca Informacijska družba - IS 2009, 12-16 Oct 2009, Ljubljana, Slovenija.

Abstract

The Pathfinder algorithm is extensively used for pruning weighted networks. It is particularly useful in the analysis of co-citation networks. The present paper reviews two versions of the algorithm. A version with improved time and space complexities, called the Binary Pathfinder, and a version optimized for sparse networks, called Sparse Pathfinder. These two algorithms were implemented and tested on text networks. The obtained results show that the algorithm optimized for sparse networks works notably faster on such data.

EPrint Type:Conference or Workshop Item (Talk)
Project Keyword:Project Keyword UNSPECIFIED
Subjects:Theory & Algorithms
ID Code:8184
Deposited By:Boris Horvat
Deposited On:21 February 2012