Biased graphwalks for RDF graph embeddings

Michael Cochez; Petar Ristoski; Simone Paolo Ponzetto; Heiko Paulheim

doi:10.1145/3102254.3102279

Biased graphwalks for RDF graph embeddings

Michael Cochez, Petar Ristoski, Simone Paolo Ponzetto, Heiko Paulheim

Research output: Chapter in Book / Report / Conference proceeding › Conference contribution › Academic › peer-review

Abstract

Knowledge Graphs have been recognized as a valuable source for background information in many data mining, information retrieval, natural language processing, and knowledge extraction tasks. However, obtaining a suitable feature vector representation from RDF graphs is a challenging task. In this paper, we extend the RDF2Vec approach, which leverages language modeling techniques for unsupervised feature extraction from sequences of entities. We generate sequences by exploiting local information from graph substructures, harvested by graph walks, and learn latent numerical representations of entities in RDF graphs. We extend the way we compute feature vector representations by comparing twelve different edge weighting functions for performing biased walks on the RDF graph, in order to generate higher quality graph embeddings. We evaluate our approach using different machine learning, as well as entity and document modeling benchmark data sets, and show that the naive RDF2Vec approach can be improved by exploiting Biased Graph Walks.

Original language	English
Title of host publication	Proceedings of the 7th International Conference on Web Intelligence, Mining and Semantics, WIMS 2017
Publisher	Association for Computing Machinery
ISBN (Electronic)	9781450352253
DOIs	https://doi.org/10.1145/3102254.3102279
Publication status	Published - 19 Jun 2017
Externally published	Yes
Event	7th International Conference on Web Intelligence, Mining and Semantics, WIMS 2017 - Amantea, Italy Duration: 19 Jun 2017 → 22 Jun 2017

Publication series

Name	ACM International Conference Proceeding Series
Volume	Part F129475

Conference

Conference	7th International Conference on Web Intelligence, Mining and Semantics, WIMS 2017
Country/Territory	Italy
City	Amantea
Period	19/06/17 → 22/06/17

Funding

Acknowledgments. Œe work presented in this paper has been partially funded by the Junior-professor funding programme of the Ministry of Science, Research and the Arts of the state of Baden-WürŠemberg (project ”Deep semantic models for high-end NLP application”), and by the German Research Foundation (DFG) under grant number PA 2373/1-1 (Mine@LOD). Œe implementation of our benchmarks was greatly aided by the work done on the Stanford Network Analysis Platform (SNAP).

Keywords

Data mining
Graph embeddings
Linked open data

Access to Document

10.1145/3102254.3102279

Persistent URL (handle)

Persistent link

Cite this

Cochez, M., Ristoski, P., Ponzetto, S. P., & Paulheim, H. (2017). Biased graphwalks for RDF graph embeddings. In Proceedings of the 7th International Conference on Web Intelligence, Mining and Semantics, WIMS 2017 Article a21 (ACM International Conference Proceeding Series; Vol. Part F129475). Association for Computing Machinery. https://doi.org/10.1145/3102254.3102279

@inproceedings{d429ccdfe6324831865a328c429667a2,

title = "Biased graphwalks for RDF graph embeddings",

abstract = "Knowledge Graphs have been recognized as a valuable source for background information in many data mining, information retrieval, natural language processing, and knowledge extraction tasks. However, obtaining a suitable feature vector representation from RDF graphs is a challenging task. In this paper, we extend the RDF2Vec approach, which leverages language modeling techniques for unsupervised feature extraction from sequences of entities. We generate sequences by exploiting local information from graph substructures, harvested by graph walks, and learn latent numerical representations of entities in RDF graphs. We extend the way we compute feature vector representations by comparing twelve different edge weighting functions for performing biased walks on the RDF graph, in order to generate higher quality graph embeddings. We evaluate our approach using different machine learning, as well as entity and document modeling benchmark data sets, and show that the naive RDF2Vec approach can be improved by exploiting Biased Graph Walks.",

keywords = "Data mining, Graph embeddings, Linked open data",

author = "Michael Cochez and Petar Ristoski and Ponzetto, {Simone Paolo} and Heiko Paulheim",

year = "2017",

month = jun,

day = "19",

doi = "10.1145/3102254.3102279",

language = "English",

series = "ACM International Conference Proceeding Series",

publisher = "Association for Computing Machinery",

booktitle = "Proceedings of the 7th International Conference on Web Intelligence, Mining and Semantics, WIMS 2017",

note = "7th International Conference on Web Intelligence, Mining and Semantics, WIMS 2017 ; Conference date: 19-06-2017 Through 22-06-2017",

}

Cochez, M, Ristoski, P, Ponzetto, SP & Paulheim, H 2017, Biased graphwalks for RDF graph embeddings. in Proceedings of the 7th International Conference on Web Intelligence, Mining and Semantics, WIMS 2017., a21, ACM International Conference Proceeding Series, vol. Part F129475, Association for Computing Machinery, 7th International Conference on Web Intelligence, Mining and Semantics, WIMS 2017, Amantea, Italy, 19/06/17. https://doi.org/10.1145/3102254.3102279

Biased graphwalks for RDF graph embeddings. / Cochez, Michael; Ristoski, Petar; Ponzetto, Simone Paolo et al.
Proceedings of the 7th International Conference on Web Intelligence, Mining and Semantics, WIMS 2017. Association for Computing Machinery, 2017. a21 (ACM International Conference Proceeding Series; Vol. Part F129475).

Research output: Chapter in Book / Report / Conference proceeding › Conference contribution › Academic › peer-review

TY - GEN

T1 - Biased graphwalks for RDF graph embeddings

AU - Cochez, Michael

AU - Ristoski, Petar

AU - Ponzetto, Simone Paolo

AU - Paulheim, Heiko

PY - 2017/6/19

Y1 - 2017/6/19

N2 - Knowledge Graphs have been recognized as a valuable source for background information in many data mining, information retrieval, natural language processing, and knowledge extraction tasks. However, obtaining a suitable feature vector representation from RDF graphs is a challenging task. In this paper, we extend the RDF2Vec approach, which leverages language modeling techniques for unsupervised feature extraction from sequences of entities. We generate sequences by exploiting local information from graph substructures, harvested by graph walks, and learn latent numerical representations of entities in RDF graphs. We extend the way we compute feature vector representations by comparing twelve different edge weighting functions for performing biased walks on the RDF graph, in order to generate higher quality graph embeddings. We evaluate our approach using different machine learning, as well as entity and document modeling benchmark data sets, and show that the naive RDF2Vec approach can be improved by exploiting Biased Graph Walks.

AB - Knowledge Graphs have been recognized as a valuable source for background information in many data mining, information retrieval, natural language processing, and knowledge extraction tasks. However, obtaining a suitable feature vector representation from RDF graphs is a challenging task. In this paper, we extend the RDF2Vec approach, which leverages language modeling techniques for unsupervised feature extraction from sequences of entities. We generate sequences by exploiting local information from graph substructures, harvested by graph walks, and learn latent numerical representations of entities in RDF graphs. We extend the way we compute feature vector representations by comparing twelve different edge weighting functions for performing biased walks on the RDF graph, in order to generate higher quality graph embeddings. We evaluate our approach using different machine learning, as well as entity and document modeling benchmark data sets, and show that the naive RDF2Vec approach can be improved by exploiting Biased Graph Walks.

KW - Data mining

KW - Graph embeddings

KW - Linked open data

UR - http://www.scopus.com/inward/record.url?scp=85028308521&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85028308521&partnerID=8YFLogxK

U2 - 10.1145/3102254.3102279

DO - 10.1145/3102254.3102279

M3 - Conference contribution

AN - SCOPUS:85028308521

T3 - ACM International Conference Proceeding Series

BT - Proceedings of the 7th International Conference on Web Intelligence, Mining and Semantics, WIMS 2017

PB - Association for Computing Machinery

T2 - 7th International Conference on Web Intelligence, Mining and Semantics, WIMS 2017

Y2 - 19 June 2017 through 22 June 2017

ER -

Biased graphwalks for RDF graph embeddings

Abstract

Publication series

Conference

Funding

Keywords

Access to Document

Persistent URL (handle)

Other files and links

Fingerprint

Cite this