Transforming Graph Data for Statistical Relational Learning

Abstract

Relational data representations have become an increasingly important topic due to the recent proliferation of network datasets (e.g., social, biological, information networks) and a corresponding increase in the application of statistical relational learning (SRL) algorithms to these domains. In this article, we examine a range of representation issues for graph-based relational data. Since the choice of relational data representation-for the nodes, links, and features-can dramatically affect the capabilities of SRL algorithms, we survey approaches and opportunities for relational representation transformation designed to improve the performance of these algorithms. This leads us to introduce an intuitive taxonomy for data representation transformations in relational domains that incorporates link transformation and node transformation as symmetric representation tasks. In particular, the transformation tasks for both nodes and links include (i) predicting their existence, (ii) predicting their label or type, (iii) estimating their weight or importance, and (iv) systematically constructing their relevant features. We motivate our taxonomy through detailed examples and use it to survey and compare competing approaches for each of these tasks. We also discuss general conditions for transforming links, nodes, and features. Finally, we highlight challenges that remain to be addressed.The majority of research in machine learning assumes independently and identically distributed data. This independence assumption is often violated in relational data, which encode dependencies among data instances. For instance, people are often linked by business associations, and information about one person can be highly informative for a prediction task involving an associate of that person. More generally, relational data can be described as a set of nodes, which can be connected by one or more types of relations (or "links"). Relational information is seemingly ubiquitous; it is present in domains such as the Internet and the worldwide web (Faloutsos et al.

Modeling dynamic behavior in large evolving graphs

“…Given a sequence of network snapshots (graphs and attributes), the Dynamic Behavioral Mixed Membership Model (DBMM) consists of (1) automatically learning a set of representative features, (2) extracting features from each graph, (3) discovering behavioral roles (4) iteratively extracting these roles from the sequence of network snapshots over time and (5) learning a predictive model of how these behaviors change over time. As an aside, let us note that DBMM is a scalable general framework for analyzing temporal behavior as the model components can be replaced by others and each component can be appropriately tuned for any application (e.g., for the feature set, any feature construction system from [25] can conceivably be used).…”

Network Structure, Metadata, and the Prediction of Missing Nodes and Annotations

“…This is done by leveraging the information available in the metadata to predict missing nodes in the network. This contrasts with the more common approach of predicting missing edges [21][22][23][24][25][26][27], which cannot be used when entire nodes have not been observed and need to be predicted, and with other approaches to detect missing nodes, which are either heuristic in nature [28], or rely on very specific assumptions on the data generating process [29,30]. Furthermore, our method is also capable of clustering the metadata themselves, separating them in equivalence classes according to their occurrence in the network.…”

A survey of graphs in natural language processing

“…Rossi et al (2012) examine and categorize techniques for transforming graph-based relational data -transformation of nodes/edges/features -to improve statistical relational learning. Rossi et al present a taxonomy for data representation transformation in relational domains that incorporates link transformation and node transformation as symmetric representation tasks.…”