The Importance of Outlier Relationships in Mobile Call Graphs

Mendiratta, Veena; Phadke, Chitra; Uzunalioğlu, Hüseyin

doi:10.1109/icmla.2012.135

Cited by 5 publications

(5 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, vertices with a very high in-degree (e.g., customer service numbers) or out-degree (e.g., salesmen) lose that correlation. Similar conclusions are drawn by Doran et al [18], although with slightly different power law parameters (an exponent of 3.41 for the in-degree and 2.63 for the out-degree).…”

Section: Mobile Call Graph Propertiessupporting

confidence: 86%

“…Instead, Doran et al [18] are only in partial agreement with the conclusion above. They rank edges according to their outlying behavior, i.e., how significantly the edge weight and overlap 20 deviate from the mean value in the graph, either positively or negatively.…”

Section: Mobile Call Graph Propertiesmentioning

confidence: 82%

“…They find the distribution of the activity rate to be heavy-tailed, with an exponent of 2.8. On the spatial side, Lambiotte et al [20] associate geographical information from billing ZIP codes to the vertices of the graph, and find that a power-law gravity model 18 well approximates the probability that two mobile users living at a given distance are connected in the graph, i.e., call each other. Assortativity.…”

Section: Mobile Call Graph Propertiesmentioning

confidence: 99%

“…Finally, Hidalgo and Rodriguez-Sickert [25] identify a relationship among the importance of vertices and edges in the mobile call graph structure. They measure the former as the vertex degree and the 18 The gravity model commends that a measure decreases as a power of distance. Denoting as d ij the geographical distance between mobile users i and j, their probability to be connected is P…”

Section: Mobile Call Graph Propertiesmentioning

confidence: 99%

See 3 more Smart Citations

Large-Scale Mobile Traffic Analysis: A Survey

Naboulsi

Fiore

Ribot

et al. 2016

IEEE Commun. Surv. Tutorials

215

161

View full text Add to dashboard Cite

This article surveys the literature on analyses of mobile traffic collected by operators within their network infrastructure. This is a recently emerged research field, and, apart from a few outliers, relevant works cover the period from 2005 to date, with a sensible densification over the last three years. We provide a thorough review of the multidisciplinary activities that rely on mobile traffic datasets, identifying major categories and sub-categories in the literature, so as to outline a hierarchical classification of research lines. When detailing the works pertaining to each class, we balance a comprehensive view of state-ofthe-art results with punctual focuses on the methodological aspects. Our approach provides a complete introductory guide to the research based on mobile traffic analysis. It allows summarizing the main findings of the current state-of-the-art, as well as pinpointing important open research directions.

show abstract

Section: Mobile Call Graph Propertiessupporting

confidence: 86%

Section: Mobile Call Graph Propertiesmentioning

confidence: 82%

Section: Mobile Call Graph Propertiesmentioning

confidence: 99%

Section: Mobile Call Graph Propertiesmentioning

confidence: 99%

See 2 more Smart Citations

Large-Scale Mobile Traffic Analysis: A Survey

Naboulsi

Fiore

Ribot

et al. 2016

IEEE Commun. Surv. Tutorials

215

161

View full text Add to dashboard Cite

show abstract

“…Figure 9 plots the distribution of these durations and exhibits a clear power-tailed shape. The power tail in the distribution of call times are a common phenomenon across mobile phone datasets [14,44,8]. We therefore only consider communication between towers whose cumulative duration of all conversations fall in the top 1.5% of this distribution, where significant calling activity occurs.…”

Section: Centralitymentioning

confidence: 99%

Operationalizing Central Place and Central Flow Theory With Mobile Phone Data

Fox

2015

Ann. Data. Sci.

Self Cite

View full text Add to dashboard Cite

Central Place and Central Flow Theory are geographic principles explaining why and how cities develop across large regional spaces. Central Place Theory postulates that cities selforganize into a spatial hierarchy were small numbers of very large 'Central Places' support numerous surrounding and less developed 'Low Places', while 'Middle Places' develop at the periphery of where Central Places carry spatial influence. Central Flow Theory is a complementary notion that explains the cooperative development of cities through joint information sharing. Both theories are often discussed, with multiple regional development and economic models built upon their tenents. However, it is very difficult to quantify the degree to which Central Place and Central Flow Theory explains the development and positions of cities in a region, particularly in developing countries where socioeconomic data is difficult to collect. To facilitate these measurements, this paper presents a way to operationalize Central Place and Central Flow Theory using mobile phone data collected across a region. It defines a set of mobile phone data attributes that are related to basic facets of the two theories, and demonstrates how their measurements speak to the degree to which the theories hold in the region the mobile phone data covers. The theory is then applied in a case study where promising locations for economic investment in a developing nation are identified.

show abstract

Propagation Models and Analysis for Mobile Phone Data Analytics

Mendiratta

2015

Intelligent Systems Reference Library

View full text Add to dashboard Cite

The Importance of Outlier Relationships in Mobile Call Graphs

Cited by 5 publications

References 11 publications

Large-Scale Mobile Traffic Analysis: A Survey

Large-Scale Mobile Traffic Analysis: A Survey

Operationalizing Central Place and Central Flow Theory With Mobile Phone Data

Propagation Models and Analysis for Mobile Phone Data Analytics

Contact Info

Product

Resources

About