Process mining classification with a weightless neural network
Authors:
Rafael Garcia Barbastefano,
Maria Clara Lippi,
Diego Carvalho
Abstract:
Using a weightless neural network architecture WiSARD we propose a straightforward graph to retina codification to represent business process graph flows avoiding kernels, and we present how WiSARD outperforms the classification performance with small training sets in the process mining context.
Using a weightless neural network architecture WiSARD we propose a straightforward graph to retina codification to represent business process graph flows avoiding kernels, and we present how WiSARD outperforms the classification performance with small training sets in the process mining context.
△ Less
Submitted 25 September, 2020;
originally announced September 2020.
Identifying influential patents in citation networks using enhanced VoteRank centrality
Authors:
João C. S. Freitas,
Rafael Barbastefano,
Diego Carvalho
Abstract:
This study proposes the usage of a method called VoteRank, created by Zhang et al. (2016), to identify influential nodes on patent citation networks. In addition, it proposes enhanced VoteRank algorithms, extending the Zhang et al. work. These novel algorithms comprise a reduction on the voting ability of the nodes affected by a chosen spreader if the nodes are distant from the spreader. One metho…
▽ More
This study proposes the usage of a method called VoteRank, created by Zhang et al. (2016), to identify influential nodes on patent citation networks. In addition, it proposes enhanced VoteRank algorithms, extending the Zhang et al. work. These novel algorithms comprise a reduction on the voting ability of the nodes affected by a chosen spreader if the nodes are distant from the spreader. One method uses a reduction factor that is linear regarding the distance from the spreader, which we called VoteRank-LRed. The other method uses a reduction factor that is exponential concerning the distance from the spreader, which we called VoteRank-XRed. By applying the methods to a citation network, we were able to demonstrate that VoteRank-LRed improved performance in choosing influence spreaders more efficiently than the original VoteRank on the tested citation network.
△ Less
Submitted 5 November, 2018;
originally announced November 2018.
Bus transport network analysis in Rio de Janeiro based on topological models using Social Networks
Authors:
Louise Pumar,
Rafael Barbastefano,
Diego Carvalho
Abstract:
In recent years, studies on public transport networks have intensified in the Social Networks field, especially in bus networks. This has occurred because of the relevance of urban mobility for the proper functioning of a city. The Rio de Janeiro city, Brazil, has undergone recent changes in the city of Rio de Janeiro, Brazil, has undergone recent changes in its municipal bus system, modifying sev…
▽ More
In recent years, studies on public transport networks have intensified in the Social Networks field, especially in bus networks. This has occurred because of the relevance of urban mobility for the proper functioning of a city. The Rio de Janeiro city, Brazil, has undergone recent changes in the city of Rio de Janeiro, Brazil, has undergone recent changes in its municipal bus system, modifying several lines and bus stops. This paper proposes to analyze the structure of the bus transportation network of this city, comparing its topology in 2014 and 2016, before and after the change. For this, the properties of the bus system were investigated based on the topological models B\-space, P\-space, and C\-space. Some essential parameters were calculated, such as the giant component, distance, diameter, degree, closeness, and betweenness. The results showed that there was a reduction of 22.75\% of the lines and 5.19\% of the bus stops from 2014 to 2016. It was also verified that a maximum of four lines is required to move between any two bus stops within the city in both years. However, with three lines is possible to reach more than 99\% of the bus stops. Besides, this study also suggests exploring the C\-space network according to a minimum number of frequent bus stops that the lines had. Based on the component giant analysis of these networks with many common points, it is possible to detect possible expressway corridors.
△ Less
Submitted 16 August, 2018;
originally announced August 2018.
