Counts-of-counts similarity for prediction and search in relational data

Manfred Jaeger, Marco Lippi, Giovanni Pellegrini, Andrea Passerini

Research output: Contribution to journalJournal articleResearchpeer-review

54 Downloads (Pure)

Abstract

Defining appropriate distance functions is a crucial aspect of effective and efficient similarity-based prediction and retrieval. Relational data are especially challenging in this regard. By viewing relational data as multi-relational graphs, one can easily see that a distance between a pair of nodes can be defined in terms of a virtually unlimited class of features, including node attributes, attributes of node neighbors, structural aspects of the node neighborhood and arbitrary combinations of these properties. In this paper we propose a rich and flexible class of metrics on graph entities based on earth mover’s distance applied to a hierarchy of complex counts-of-counts statistics. We further propose an approximate version of the distance using sums of marginal earth mover’s distances. We show that the approximation is correct for many cases of practical interest and allows efficient nearest-neighbor retrieval when combined with a simple metric tree data structure. An experimental evaluation on two real-world scenarios highlights the flexibility of our framework for designing metrics representing different notions of similarity. Substantial improvements in similarity-based prediction are reported when compared to solutions based on state-of-the-art graph kernels.
Original languageEnglish
JournalData Mining and Knowledge Discovery
Volume33
Issue number5
Pages (from-to)1254–1297
Number of pages44
ISSN1384-5810
DOIs
Publication statusPublished - Sep 2019

Keywords

  • Earth-mover’s distance
  • Graph mininga
  • Relational data mining
  • Similarity search
  • Statistical relational learning

Fingerprint

Dive into the research topics of 'Counts-of-counts similarity for prediction and search in relational data'. Together they form a unique fingerprint.

Cite this