From Eigentrust to a Trust-measuring Algorithm in the Max-Plus Algebra

Juan Afanador , Nir Oren, Murilo Baptista, Maria Araujo

Research output: Contribution to conferencePaper

Abstract

Eigentrust has become a popular method for trust computation based on both direct and indirect interaction information. Eigentrust computes a final trust value based on eigenvectors computed via the Power Method. However, under certain network topologies, the Power Method cannot be used to identify appropriate eigenvectors. After characterising these cases, we overcome Eigentrust’s limitations by extending the algorithm’s core ideas into the Max-Plus algebra. An empirical evaluation of our new approach demonstrates its superiority to Eigentrust.
Original languageEnglish
Publication statusAccepted/In press - 14 Jan 2020
Event24th European Conference on Artificial Intelligence - Santiago de Compostela, Spain
Duration: 8 Jun 202012 Jun 2020
http://ecai2020.eu/

Conference

Conference24th European Conference on Artificial Intelligence
Abbreviated titleECAI 2020
CountrySpain
CitySantiago de Compostela
Period8/06/2012/06/20
Internet address

Fingerprint

Eigenvalues and eigenfunctions
Algebra
Topology

Cite this

Afanador , J., Oren, N., Baptista, M., & Araujo, M. (Accepted/In press). From Eigentrust to a Trust-measuring Algorithm in the Max-Plus Algebra. Paper presented at 24th European Conference on Artificial Intelligence , Santiago de Compostela, Spain.

From Eigentrust to a Trust-measuring Algorithm in the Max-Plus Algebra. / Afanador , Juan; Oren, Nir; Baptista, Murilo; Araujo, Maria.

2020. Paper presented at 24th European Conference on Artificial Intelligence , Santiago de Compostela, Spain.

Research output: Contribution to conferencePaper

Afanador , J, Oren, N, Baptista, M & Araujo, M 2020, 'From Eigentrust to a Trust-measuring Algorithm in the Max-Plus Algebra', Paper presented at 24th European Conference on Artificial Intelligence , Santiago de Compostela, Spain, 8/06/20 - 12/06/20.
Afanador J, Oren N, Baptista M, Araujo M. From Eigentrust to a Trust-measuring Algorithm in the Max-Plus Algebra. 2020. Paper presented at 24th European Conference on Artificial Intelligence , Santiago de Compostela, Spain.
Afanador , Juan ; Oren, Nir ; Baptista, Murilo ; Araujo, Maria. / From Eigentrust to a Trust-measuring Algorithm in the Max-Plus Algebra. Paper presented at 24th European Conference on Artificial Intelligence , Santiago de Compostela, Spain.
@conference{0bdb869928644a9c9877db399ef9558c,
title = "From Eigentrust to a Trust-measuring Algorithm in the Max-Plus Algebra",
abstract = "Eigentrust has become a popular method for trust computation based on both direct and indirect interaction information. Eigentrust computes a final trust value based on eigenvectors computed via the Power Method. However, under certain network topologies, the Power Method cannot be used to identify appropriate eigenvectors. After characterising these cases, we overcome Eigentrust’s limitations by extending the algorithm’s core ideas into the Max-Plus algebra. An empirical evaluation of our new approach demonstrates its superiority to Eigentrust.",
author = "Juan Afanador and Nir Oren and Murilo Baptista and Maria Araujo",
note = "The proceedings of the conference will be published by IOS Press and made freely available at the IOS website (https://www.iospress.nl/?s=ECAI); 24th European Conference on Artificial Intelligence , ECAI 2020 ; Conference date: 08-06-2020 Through 12-06-2020",
year = "2020",
month = "1",
day = "14",
language = "English",
url = "http://ecai2020.eu/",

}

TY - CONF

T1 - From Eigentrust to a Trust-measuring Algorithm in the Max-Plus Algebra

AU - Afanador , Juan

AU - Oren, Nir

AU - Baptista, Murilo

AU - Araujo, Maria

N1 - The proceedings of the conference will be published by IOS Press and made freely available at the IOS website (https://www.iospress.nl/?s=ECAI)

PY - 2020/1/14

Y1 - 2020/1/14

N2 - Eigentrust has become a popular method for trust computation based on both direct and indirect interaction information. Eigentrust computes a final trust value based on eigenvectors computed via the Power Method. However, under certain network topologies, the Power Method cannot be used to identify appropriate eigenvectors. After characterising these cases, we overcome Eigentrust’s limitations by extending the algorithm’s core ideas into the Max-Plus algebra. An empirical evaluation of our new approach demonstrates its superiority to Eigentrust.

AB - Eigentrust has become a popular method for trust computation based on both direct and indirect interaction information. Eigentrust computes a final trust value based on eigenvectors computed via the Power Method. However, under certain network topologies, the Power Method cannot be used to identify appropriate eigenvectors. After characterising these cases, we overcome Eigentrust’s limitations by extending the algorithm’s core ideas into the Max-Plus algebra. An empirical evaluation of our new approach demonstrates its superiority to Eigentrust.

M3 - Paper

ER -