Database Paper Browser

Back to papers

MINT: Detecting Fraudulent Behaviors from Time-series Relational Data

Summary: MINT constructs a time-aware behavior graph per user (rows as action nodes) and learns hierarchical short/medium/long-term intentions via three temporal GCNs with a gated neighbor interaction to capture row-level effects and avoid over-smoothing. Its exponential receptive-field design uses far fewer GCN layers (no RNN), improving training efficiency and scalability on billion-scale e-commerce logs and outperforming 10 SOTA models with better interpretability. (summarized by gpt-5-mini on Feb 09 2026)

Paper ID
13191
Venue
VLDB
Year
2023
Pagerank
4.1945683e-05
Overall Rank
11,266 | 21.63%
DOI
10.14778/3611540.3611551

Incoming Non-self Citations Over Time

No non-self incoming citations found for this paper in this database.

Authors

Incoming Citations (Sorted by Pagerank)

Showing 0 of 0 citing papers.

Rank Citing Paper Year Venue Pagerank
Previous Page 1 / 1 Next

Outgoing Citations (Sorted by Pagerank)

Showing 12 of 12 cited papers.

Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.

Previous Page 1 / 1 Next

Semantically Similar Papers