Enhanced Customer Churn Prediction using Social Network Analysis

Abd-Allah, Marwa; Salah, Akram; El-Beltagy, Samhaa R.

doi:10.1145/2665994.2665997

Cited by 7 publications

(5 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In Abd-Allah et al [32], they constructed an undirected call graph and evaluated the social tie strength using new interactional attributes, derived from basic call attributes, to ensure the validity of the calculated tie strength. They proposed an influence propagation model, where the strongest ties are exploited for churn influence transfers from one node to another.…”

Section: Social Churning Predictionmentioning

confidence: 99%

A social model based on customers’ profiles for analyzing the churning process in the mobile market of data plans

Postigo-Boix

Melús-Moreno

2018

Physica A: Statistical Mechanics and its Applications

View full text Add to dashboard Cite

Mobile Network Operators (MNOs) present wireless services of the same kind in identical zones, clients select the service taking into account any element they consider relevant. Churning hits on the design of the network and the method to assign prices by MNOs, and of course their earnings. Therefore, MNOs try to reduce churn detecting potential churners before they leave the service. Our approach to churn prediction considers each customer individually. Previous research shows that members of the social circle of a subscriber may influence churn. Thus, many scenarios that describe social relations, and in which churning processes could be expected, set an emerging challenge with practical implications. This paper uses the Agent-Based Modelling (ABM) technique to model customers. The model’s parameters include demographic and psychographic features as well as usage profiles according to their social behavior considering their customers’ profiles. Our model modifies and extends an existing real social network generator algorithm that considers customer’s profiles and homophily considerations to create connections. We show that using our approach, groups of customers with greater tendency to churn due to the influence of their social networks can be identified better.Peer ReviewedPostprint (author's final draft

show abstract

Section: Social Churning Predictionmentioning

confidence: 99%

A social model based on customers’ profiles for analyzing the churning process in the mobile market of data plans

Postigo-Boix

Melús-Moreno

2018

Physica A: Statistical Mechanics and its Applications

View full text Add to dashboard Cite

show abstract

“…Concerning churn prediction, these works use different approaches depending on the dataset. There are two approaches to detect churns: one non-relational (classical approach) [15], [16] and the another one relational (social approach) [17], [6], [18], [8], [9]. We describe these approaches in the following subsections.…”

Section: Related Workmentioning

confidence: 99%

Measuring churner influence on pre-paid subscribers using fuzzy logic

Columelli

Núñez-del-Prado

Zarate-Gamarra³

2016

2016 XLII Latin American Computing Conference (CLEI)

View full text Add to dashboard Cite

In the last decades, mobile phones have become the major medium for communication between humans. The site effect is the loss of subscribers. Consequently, Telecoms operators invest in developing algorithms for quantifying the risk to churn and to influence other subscribers to churn. The objective is to prioritize the retention of subscribers in their network due to the cost of obtaining a new subscriber is four times more expensive than retaining subscribers. Hence, we use Extremely Random Forest to classify churners and non-churners obtaining a Lift value at 10% of 5.5. Then, we rely on graph-based measures such as Degree of Centrality and Page rank to measure emitted and received influence in the social network of the carrier. Our methodology allows summarising churn risk score, relying on a Fuzzy Logic system, combining the churn probability and the risk of the churner to leave the network with other subscribers.

show abstract

“…These techniques have long been known and used in many fields, however, as indicated by a number of authors, the use of one of them does not guarantee to sufficiently exploit the information hidden in the telecoms operation data [6]- [8]. Application of social networks analysis [9], [10] is proposed by many authors [11], [12], but usually it is employed to analyze the relationship between subscribers only [13]- [15]. In the method proposed in this paper both kinds of data are used: common statistics of traffic data, as well as those resulting from social network analysis.…”

Section: Similar Solutions and Related Workmentioning

confidence: 99%

Application of Social Network Inferred Data to Churn Modeling in Telecoms

Gruszczyński

2016

JTIT

View full text Add to dashboard Cite

The subject of this work is the use of social network analysis to increase the effectiveness of methods used to predict churn of telephony network subscribers. The social network is created on the basis of operational data (CDR records). The result of the analysis is customer segmentation and additional predictor variables. Proposed hybrid predictor employs set of regression models tuned to specific customer segments. The verification was performed on data obtained from one of the Polish operators.

show abstract

Enhanced Customer Churn Prediction using Social Network Analysis

Cited by 7 publications

References 9 publications

A social model based on customers’ profiles for analyzing the churning process in the mobile market of data plans

A social model based on customers’ profiles for analyzing the churning process in the mobile market of data plans

Measuring churner influence on pre-paid subscribers using fuzzy logic

Application of Social Network Inferred Data to Churn Modeling in Telecoms

Contact Info

Product

Resources

About