Link prediction using time series of neighborhood-based node similarity scores

Gunes, Ismail; Gündüz-Ögüdücü, Şule; Çataltepe, Zehra

doi:10.1007/s10618-015-0407-0

Cited by 96 publications

(61 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…1) Univariate Time Series Models: Perhaps the most straightforward approach to dynamic link prediction is to apply standard univariate time series models to each node pair. Autoregressive Integrated Moving Average (ARIMA) models were used for dynamic link prediction in studies [7], [8]. A special case, the ARIMA(0, 1, 0) model, is an exponentiallyweighted moving average (EWMA) model, which has been used in studies [4], [6], [14], [15].…”

Section: Methods For Dynamic Link Predictionmentioning

confidence: 99%

“…Huang and Lin [8] aggregated the dynamic network over time to form a static network then apply similarity-based methods. Güneş et al [7] computed node similarities at each time step then model these similarities using ARIMA models. Dunlavy et al [14] proposed a truncated version of the Katz predictor based on a low-rank approximation of a weighted average of past adjacency matrices.…”

Section: Methods For Dynamic Link Predictionmentioning

confidence: 99%

“…For instance, in a social network with timestamped edges denoting interactions between people (nodes), an edge may appear at several time instances where a pair of people are frequently interacting then disappear after interactions cease. Since existing edges between nodes may be [6] 0.780 0.239 0.371 TS-AA [7] 0.777 0.065 0.144 TS-Katz [8] 0.879 0.077 0.149 SBTM [9] 0.799 0.138 0.337 removed at a future time, the dynamic link prediction problem is more complex and also involves computing a predicted score for existing edges, because they may disappear at a future time.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Evaluating Link Prediction Accuracy in Dynamic Networks with Added and Removed Edges

Junuthula

Devabhaktuni

2016

2016 IEEE International Conferences on Big Data and Cloud Computing (BDCloud), Social Computing and Networking (SocialCom), Sus

View full text Add to dashboard Cite

In this paper, we investigate several metrics currently used for evaluating accuracies of dynamic link prediction methods and demonstrate why they can be misleading in many cases. We provide several recommendations on evaluating dynamic link prediction accuracy, including separation into two categories of evaluation. Finally we propose a unified metric to characterize link prediction accuracy effectively using a single number.

show abstract

Section: Methods For Dynamic Link Predictionmentioning

confidence: 99%

Section: Methods For Dynamic Link Predictionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Evaluating Link Prediction Accuracy in Dynamic Networks with Added and Removed Edges

Junuthula

Devabhaktuni

2016

2016 IEEE International Conferences on Big Data and Cloud Computing (BDCloud), Social Computing and Networking (SocialCom), Sus

View full text Add to dashboard Cite

show abstract

“…The major limitation of this work is its inability to predict new links, that is, the method only predicts re-occurring links. This limitation was addressed in precedent works by considering other types of features such as similarity scores (da Silva Soares & Prudêncio, 2012;Güneş, Gündüz-Ö güdücü, & Çataltepe, 2016;Hajibagheri, Sukthankar, & Lakkaraju, 2016). These methods also extended the existing work by considering other forecasting models.…”

Section: Link Predictionmentioning

confidence: 99%

Social network analysis: An overview

Tabassum

Pereira

Fernandes

et al. 2018

WIREs Data Min & Knowl

244

129

View full text Add to dashboard Cite

Social network analysis (SNA) is a core pursuit of analyzing social networks today. In addition to the usual statistical techniques of data analysis, these networks are investigated using SNA measures. It helps in understanding the dependencies between social entities in the data, characterizing their behaviors and their effect on the network as a whole and over time. Therefore, this article attempts to provide a succinct overview of SNA in diverse topological networks (static, temporal, and evolving networks) and perspective (ego‐networks). As one of the primary applicability of SNA is in networked data mining, we provide a brief overview of network mining models as well; by this, we present the readers with a concise guided tour from analysis to mining of networks. This article is categorized under: Application Areas > Science and Technology Technologies > Machine Learning Fundamental Concepts of Data and Knowledge > Human Centricity and User Interaction Commercial, Legal, and Ethical Issues > Social Considerations

show abstract

“…link prediction methods based on time-series (HUANG; LIN, 2009;SOARES;PRUDêNCIO, 2012;GÜNEŞ;GÜNDÜZ-ÖGÜDÜCÜ;ÇATALTEPE, 2016;MORADABADI;MEYBODI, 2017) and link prediction methods in dynamic networks (SARKAR; CHAKRABARTI; JORDAN, 2012;STEEG;GALSTYAN, 2014;RAHMAN;HASAN, 2016).…”

Section: Link Prediction Methodsmentioning

confidence: 99%

Mining user behavior in location-based social networks

Rebaza¹

View full text Add to dashboard Cite

Link prediction using time series of neighborhood-based node similarity scores

Cited by 96 publications

References 34 publications

Evaluating Link Prediction Accuracy in Dynamic Networks with Added and Removed Edges

Evaluating Link Prediction Accuracy in Dynamic Networks with Added and Removed Edges

Social network analysis: An overview

Mining user behavior in location-based social networks

Contact Info

Product

Resources

About