Classification in Networked Data: A Toolkit and a Univariate Case Study

By Sofus A. Macskassy, Foster J. Provost. 2007

531 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Graph Learning

Citations in the corpus

DeepWalk: online learning of social representations (2014, background): “It has shown surprisingly good performance in real networks, and has been advocated as a sensible relational classification baseline [26].”
Enhancing Transparency and Control When Drawing Data-Driven Inferences About Individuals (2016, background): “As digital data become increasingly centered on the inherent network structure of many online platforms and online inferences are incorporating social network data into their models, expanding cloaking to utilize network-based techniques can have a dramatic effect on inference and cloakability.(31) ”
Feature Selection for Classification: A Review (2014, background): “Many linked data related learning tasks are proposed such as collective classification [47, 59], and relational clustering [44, 45], but the task of feature selection for linked data is rarely touched.”
A bias/variance decomposition for models using collective inference (2008, background): “relational models that do not learn at all but simply assume autocorrelation is present while propagating information throughout the graph for inference (e.g., Macskassy and Provost 2007 ).”
Graph Regularized Transductive Classification on Heterogeneous Information Networks (2010, methodology): “– Learning with Local and Global Consistency (LLGC) [21] – Weighted-vote Relational Neighbor classifier (wvRN) [10] [11] – Network-only Link-based classification (nLB) [9] [11]”
Collective Classi!cation in Network Data (2008, background): “In addition, Macskassy and Provost (2007) have investigated approaches that only make use of the labels of neighbors.”
Collective Data-Sanitization for Preventing Sensitive Information Inference Attacks in Social Networks (2018, methodology): “To prevent this, the weighted-vote Relational Neighbor algorithm (wvRN) [28] suggests to add a weight to each friendship link.”
Local Modeling of Attributed Graphs: Algorithms and Applications (2016, background): “It has shown surprisingly good performance in real networks, and has been advocated as a sensible relational classification baseline [84].”
Recommender Systems (2016, unknown): “A toolkit for collective classification may be found in [388].”
Leveraging social media networks for classification (2011, methodology): “WVrn has been shown to work reasonably well for classification in the univariate case and is recommended as a baseline method for comparison [29].”
Collective Classification in Network Data (2008, background): “In addition, Macskassy and Provost (2007) have investigated approaches that only make use of the labels of neighbors.”
Community Detection and Mining in Social Media (2010, methodology): “In order to address the classification task with network data, we typically need three components (Macskassy and Provost, 2007): a local classifier, a relational classifier and collective inference.”