Clustering and predicting the data usage patterns of geographically diverse mobile users

Walelgne, Ermias Andargie; Asrese, Alemnew Sheferaw; Manner, Jukka; Bajpai, Vaibhav; Ott, Jörg

doi:10.1016/j.comnet.2020.107737

Cited by 6 publications

(5 citation statements)

References 34 publications

(31 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Heavy users, comprising only about 3% of the total, used more data than approximately 97% of the other users. A similar phenomenon was found in other countries, although the proportions were different [14]. Furthermore, similar trends were found in a study that grouped users by analyzing actual OTT traffic.…”

Section: Problem Statement 231 Class-imbalance Problemsupporting

confidence: 84%

“…While these studies are significant in that they dealt with ML methods specialized for OTT traffic classification, they neglected real-world problems that were common to the OTT market. OTT consumers range from heavy users who watch tremendous amounts of OTT content, to light users who rarely watch, and the proportions are not the same [12,14]. If ML is applied without considering these imbalances, the classification accuracy drops for the smaller user group [15].…”

Section: Motivation and Objectivementioning

confidence: 99%

“…As such, OTT market consumers can be divided into various groups, and the number of group members varies widely. According to Walelgne et al [14], who analyzed the data traffic of mobile users (including OTT), heavy users who comprise only 2-4% of the total consume the most data. They collected mobile traffic from Finland, Germany, the UK, Japan, and Brazil and classified users using a clustering method.…”

Section: Problem Statement 231 Class-imbalance Problemmentioning

confidence: 99%

See 2 more Smart Citations

A Heterogeneous Learning Framework for Over-the-Top Consumer Analysis Reflecting the Actual Market Environment

Choi

Kim

2021

Applied Sciences

View full text Add to dashboard Cite

The over-the-top (OTT) market for media consumption over wired and wireless Internet is growing. It is, therefore, crucial that service providers and carriers participating in the OTT market analyze consumer traffic for pricing, service delivery, infrastructure investments, etc. The OTT market has many consumer groups, but the proportion of users is not consistent in each. Furthermore, as multimedia consumption has increased owing to the COVID-19 epidemic, the OTT market has changed rapidly. If this is not reflected, the analysis will not be accurate. Therefore, we propose a framework that can classify consumers well based on actual OTT market environment conditions. First, by applying our proposed conditional probability-based method to basic machine learning techniques, such as support vector machine, k-nearest neighbor, and decision tree, we can improve the classification performance, even for an imbalanced OTT consumer distribution. Then, it is possible to analyze the changing consumer trends by dynamically retraining the incoming OTT consumer data. Conventional methods result in low classification accuracy in low-number classes, but our method shows an improvement of 5.3–19.2% based on recall. Moreover, conventional methods have shown large fluctuations in performance as the OTT market environment has changed, but our framework consistently maintains high performance.

show abstract

Section: Problem Statement 231 Class-imbalance Problemsupporting

confidence: 84%

Section: Motivation and Objectivementioning

confidence: 99%

Section: Problem Statement 231 Class-imbalance Problemmentioning

confidence: 99%

See 1 more Smart Citation

A Heterogeneous Learning Framework for Over-the-Top Consumer Analysis Reflecting the Actual Market Environment

Choi

Kim

2021

Applied Sciences

View full text Add to dashboard Cite

show abstract

“…As a result, they created MobiPupose, a system that could track app network requests and classify data collection purposes based on app traffic patterns. Walelgne et al [205] and Okic et al [206] showed that different app categories have different traffic patterns. In terms of traffic volume, entertainment and social media apps are the most traffic-intensive, while education and weather apps are the least traffic-intensive [205].…”

Section: B App Traffic Patternsmentioning

confidence: 99%

“…Walelgne et al [205] and Okic et al [206] showed that different app categories have different traffic patterns. In terms of traffic volume, entertainment and social media apps are the most traffic-intensive, while education and weather apps are the least traffic-intensive [205]. Moreover, in terms of temporal patterns, music and shopping apps see the most traffic in the morning, while e-mail and game apps see the most traffic during lunchtime [206].…”

Section: B App Traffic Patternsmentioning

confidence: 99%

Smartphone App Usage Analysis: Datasets, Methods, and Applications

Tong

Wang

et al. 2022

IEEE Commun. Surv. Tutorials

View full text Add to dashboard Cite

As smartphones have become indispensable personal devices, the number of smartphone users has increased dramatically over the last decade. These personal devices, which are supported by a variety of smartphone apps, allow people to access Internet services in a convenient and ubiquitous manner. App developers and service providers can collect fine-grained app usage traces, revealing connections between users, apps, and smartphones. We present a comprehensive review of the most recent research on smartphone app usage analysis in this survey. Our survey summarizes advanced technologies and key patterns in smartphone app usage behaviors, all of which have significant implications for all relevant stakeholders, including academia and industry. We begin by describing four data collection methods: surveys, monitoring apps, network operators, and app stores, as well as nine publicly available app usage datasets. We then systematically summarize the related studies of app usage analysis in three domains: app domain, user domain, and smartphone domain. We make a detailed taxonomy of the problem studied, the datasets used, the methods used, and the significant results obtained in each domain. Finally, we discuss future directions in this exciting field by highlighting research challenges.

show abstract