Ranking-based classification of heterogeneous information networks

Abstract

It has been recently recognized that heterogeneous information networks composed of multiple types of nodes and links are prevalent in the real world. Both classification and ranking of the nodes (or data objects) in such networks are essential for network analysis. However, so far these approaches have generally been performed separately. In this paper, we combine ranking and classification in order to perform more accurate analysis of a heterogeneous information network. Our intuition is that highly ranked objects within a class should play more important roles in classification. On the other hand, class membership information is important for determining a quality ranking over a dataset. We believe it is therefore beneficial to integrate classification and ranking in a simultaneous, mutually enhancing process, and to this end, propose a novel ranking-based iterative classification framework, called RankClass. Specifically, we build a graph-based ranking model to iteratively compute the ranking distribution of the objects within each class. At each iteration, according to the current ranking results, the graph structure used in the ranking algorithm is adjusted so that the subnetwork corresponding to the specific class is emphasized, while the rest of the network is weakened. As our experiments show, integrating ranking with classification not only generates more accurate classes than the state-of-art classification methods on networked data, but also provides meaningful ranking of objects within each class, serving as a more informative view of the data than traditional classification.

A Feature-Enhanced Ranking-Based Classifier for Multimodal Data and Heterogeneous Information Networks

“…NetClus [10] used a star schema to represent heterogeneous data and considered ranking information within each object type. More recently, Ji et al [8] proposed RankClass, which performs ranking and classification at the same time. Ranking and classification enhance each other in RankClass.…”

“…We follow the terminologies from RankClass [8]. A heterogeneous information network consists of multiple types of data objects and links connecting different types of data objects.…”

“…2) Varying Cross-Type Influence λ: In RankClass, the parameter λ ij , i, j ≤ m controls how much information flows from type i nodes to type j nodes, thus affecting the ranking distribution P (x ip |X i , k) in RankClass [8] as in Eq. 2.…”

“…RankClass performs ranking and classification at the same time on heterogeneous information networks. FRankClass extends RankClass [8] to a more general framework that is applicable to most classification problems in general. In this paper, we construct heterogeneous information networks from multimodal documents and perform classification according to network structures.…”

Hybrid social media network

“…The heterogeneous information network [10,12] in the data mining community also attempts to model the relations between heterogeneous entities. However, the modeling is performed by connecting any two entities with a single kind of relation, which actually falls in the single edge graph and thus differs from our proposal here.…”