We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.SI

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Social and Information Networks

Title: Is a Single Embedding Enough? Learning Node Representations that Capture Multiple Social Contexts

Abstract: Recent interest in graph embedding methods has focused on learning a single representation for each node in the graph. But can nodes really be best described by a single vector representation? In this work, we propose a method for learning multiple representations of the nodes in a graph (e.g., the users of a social network). Based on a principled decomposition of the ego-network, each representation encodes the role of the node in a different local community in which the nodes participate. These representations allow for improved reconstruction of the nuanced relationships that occur in the graph -- a phenomenon that we illustrate through state-of-the-art results on link prediction tasks on a variety of graphs, reducing the error by up to $90\%$. In addition, we show that these embeddings allow for effective visual analysis of the learned community structure.
Subjects: Social and Information Networks (cs.SI); Machine Learning (cs.LG); Machine Learning (stat.ML)
ACM classes: I.2.6; H.2.8; G.2.2
Journal reference: In Proceedings of "The Web Conference" 2019, WWW, 2019
DOI: 10.1145/3308558.3313660
Cite as: arXiv:1905.02138 [cs.SI]
  (or arXiv:1905.02138v1 [cs.SI] for this version)

Submission history

From: Alessandro Epasto [view email]
[v1] Mon, 6 May 2019 16:46:38 GMT (7486kb,D)

Link back to: arXiv, form interface, contact.