Dirichlet PageRank and Ranking Algorithms Based on Trust and Distrust

  title={Dirichlet PageRank and Ranking Algorithms Based on Trust and Distrust},
  author={Fan Chung Graham and Alexander Tsiatas and Wensong Xu},
  journal={Internet Mathematics},
  pages={113 - 134}
Motivated by numerous models of representing trust and distrust within a network ranking system, we examine a quantitative vertex ranking with consideration of the influence of a subset of nodes. We propose and analyze a general ranking metric, called Dirichlet PageRank, which gives a ranking of vertices in a subset S of nodes subject to some specified conditions on the vertex boundary of S. In addition to the usual Dirichlet boundary condition (which disregards the influence of nodes outside… 
Generalized Preference Learning for Trust Network Inference
This work introduces the new problem of Trust Network Inference (TNI), that is, inferring a trust network from a sequence of timestamped interaction networks and proposes a principled approach based on a preference learning paradigm, under a preference-based racing formulation.
SPONGE: A generalized eigenproblem for clustering signed networks
We introduce a principled and theoretically sound spectral method for k-way clustering in signed graphs, where the affinity measure between nodes takes either positive or negative values. Our
Mining Signed Networks: Theory and Applications: Tutorial proposal for the Web Conference 2020
This tutorial aims to provide an overview of the literature in mining signed networks, the most important theoretical results since their inception to the present day, some of the most common applications, and to reflect on emerging applications and directions for future work.
A Survey of Signed Network Mining in Social Media
A review of mining signed networks in the context of social media and discuss some promising research directions and new frontiers of signed network mining.
An MBO scheme for clustering and semi-supervised clustering of signed networks
We introduce a principled method for the signed clustering problem, where the goal is to partition a graph whose edge weights take both positive and negative values, such that edges within the same
To Trust or Not to Trust Lurkers?: Evaluation of Lurking and Trustworthiness in Ranking Problems
Understanding and quantifying relations between lurkers and trustworthy/untrustworthy users in ranking problems and results obtained indicate that lurkers should not be a-priori flagged as untrustworthy users, and that trustworthy users can indeed be found among lurkers.
Predictable Dynamics of Opinion Forming for Networks With Antagonistic Interactions
A significant class of signed graphs is identified for which the linear dynamics are however predictable and many analogies with positive dynamical systems are shown, including those of adjacency matrices that are eventually positive.
Improving sybil detection in online social networks
Due to open and anonymous nature, online social networks are particularly vulnerable to Sybil attack, where a malicious user can fabricate many dummy identities to target systems. Recently, there are
Spectral Clustering of Signed Graphs via Matrix Power Means
A thorough analysis of the proposed approach in the setting of a general Stochastic Block Model shows that in expectation the signed power mean Laplacian captures the ground truth clusters under reasonable settings where state-of-the-art approaches fail.
Node Classification for Signed Social Networks Using Diffuse Interface Methods
It is shown that blending the information from both positive and negative interactions leads to performance improvement in real signed social networks, consistently outperforming the current state of the art.


A Novel Approach to Propagating Distrust
This work proposes a novel and simple extension of the Page Rank equations, and argues that it naturally captures most types of distrust that are expressed in such networks.
PageRank as a discrete Green ’ s function
Originally, PageRank was a way to assign quantitative ranking to webpages by Web search engines. In general, PageRank can be viewed as a measure of relative “importance” defined on any given graph.
The PageTrust Algorithm: How to rank web pages when negative links are allowed?
The paper introduces a novel algorithm derived from the PageRank algorithm of Brin and Page that preserves good properties of robustness against possible spammers and includes negative links and converges to a trust value for each page.
Propagation of trust and distrust
It is shown that a small number of expressed trusts/distrust per individual allows us to predict trust between any two people in the system with high accuracy.
A Sharp PageRank Algorithm with Applications to Edge Ranking and Graph Sparsification
We give an improved algorithm for computing personalized PageRank vectors with tight error bounds which can be as small as Ω(n − p ) for any fixed positive integer p. The improved PageRank algorithm
Trust-based recommendation systems: an axiomatic approach
Which of these networks that represent trust and recommendation systems that incorporate these trust relationships are incentive compatible are determined, meaning that groups of agents interested in manipulating recommendations can not induce others to share their opinion by lying about their votes or modifying their trust links.
Local Graph Partitioning using PageRank Vectors
An improved algorithm for computing approximate PageRank vectors, which allows us to find a cut with conductance at most oslash and approximately optimal balance in time O(m log4 m/oslash) in time proportional to its size.
Topic-Sensitive PageRank: A Context-Sensitive Ranking Algorithm for Web Search
It is shown that using linear combinations of these (precomputed) biased PageRank vectors to generate context-specific importance scores for pages at query time, can generate more accurate rankings than with a single, generic PageRank vector.
Pagerank Increase under Different Collusion Topologies
It is shown that highly-ranked Web sites do not benefit that much from collusion due to the power law distribution of Pagerank, which depends both on the reset probability of the random walk e and on the original Pageranks of the colluding set.
Combating Web Spam with TrustRank