Graph Ranking Auditing: Problem Definition and Fast Solutions
| Autor: | Nan Cao, Jian Kang, Yinglong Xia, Meijia Wang, Hanghang Tong, Wei Fan |
|---|---|
| Rok vydání: | 2021 |
| Předmět: |
Theoretical computer science
Computer science Approximation algorithm Graph theory 02 engineering and technology Recommender system Computer Science Applications law.invention Data modeling Computational Theory and Mathematics Ranking PageRank law 020204 information systems Scalability 0202 electrical engineering electronic engineering information engineering MathematicsofComputing_DISCRETEMATHEMATICS Information Systems Team management |
| Zdroj: | IEEE Transactions on Knowledge and Data Engineering. 33:3366-3380 |
| ISSN: | 2326-3865 1041-4347 |
| Popis: | Ranking on graphs is a centerpiece in many high-impact application domains, such as information retrieval, recommender systems, team management, neuroscience and many more. PageRank, along with many of its variants, is widely used across these application domains thanks to its mathematical elegance and the superior performance. Although PageRank and its variants are effective in ranking nodes on graphs, they often lack an efficient and effective way to audit the ranking results in terms of the input graph structure, e.g., which node or edge in the graph contributes most to the top-1 ranked node; which subgraph plays a crucial role in generating the overall ranking result? In this paper, we propose to audit graph ranking by finding the influential graph elements (e.g., edges, nodes, attributes, and subgraphs) regarding their impact on the ranking results. First, we formulate graph ranking auditing problem as quantifying the influence of graph elements on the ranking results. Second, we show that our formulation can be applied to a variety of graph structures. Third, we propose effective and efficient algorithms to find the top- $k$ k influential edges/nodes/subgraph. Finally, we perform extensive empirical evaluations on real-world datasets to demonstrate that the proposed methods ( Aurora ) provide intuitive auditing results with linear scalability. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|