Mobile Phone Usage Data for Credit Scoring

Ots, Henri; Liiv, Innar; Tur, Diana

doi:10.1007/978-3-030-57672-1_7

Cited by 8 publications

(8 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…With further details explained in section 4, studies used usage history data for identifying user or device such as by using: web browsing behavior [7], call-log [8,9], application behavior [10], and the set of apps installed [11,12]. Meanwhile, user profiling using this data is such as by using call-log [22,25] and the set of apps installed [26,27].…”

Section: Smartphone Data Taxonomymentioning

confidence: 99%

“…Smartphone usage data also can be used for user profiling (often referred to as "user identification traits" or "user fingerprinting"). Some studies, such as proposed in [38][39][40][41] revealed the communal fingerprint while others, such as the examples in [22][23][24][25][26][27] revealed the individual fingerprint from the dataset. The communal fingerprint is the general trend of the people in the dataset, such as daily activity pattern within people in the same work area, land use pattern in some different areas, or people movement pattern.…”

Section: User Profilingmentioning

confidence: 99%

“…Similar to the other two first mentioned studies, this study also used SVM as the model to predict the classification label Y from the feature vectors X. Using a dataset from an anonymous European consumer lending company offering a digital loan application submitted by a mobile application, the study in [25] examined the dataset of 2.503 customers having a loan. From the dataset, they separated the trustworthy customer (1.516 customers without any debts that exceeded the 90-days limit) and the untrustworthy customer (987 customers who delayed their payment).…”

Section: User Profilingmentioning

confidence: 99%

“…by using usage data [7][8][9][10][11][12], sensor data [13][14][15][16], and user input data [17][18][19][20][21]. User profiling, meanwhile generates users' profiles or traits (such as age, gender, income, and personality) by exploring how data are correlated with user personal information to extract key features and describe their characteristics, for instance in [22][23][24][25][26][27][28]. As any information capable of identifying or generating information about a person is categorized as 'personal data' [29], smartphone data that can identify the user in user identification and generate user information in user profiling should be started to legally entitled to the protection from exploitation and leakage.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

A review on smartphone usage data for user identification and user profiling

Auliya

Nugroho

Setiawan

2021

CST

View full text Add to dashboard Cite

The amount of retrievable smartphone data is escalating; while some apps on the smartphone are evidently exploiting and leaking users’ data. These phenomena potentially violate privacy and personal data protection laws as various studies have showed that technologies such as artificial intelligence could transform smartphone data into personal data by generating user identification and user profiling. User identification identifies specific users among the data based upon the users’ characteristics and users profiling generates users’ traits (e.g. age and personality) by exploring how data is correlated with personal information. Nevertheless, the comprehensive review papers discussing both of the topics are limited. This paper thus aims to provide a comprehensive review of user identification and user profiling using smartphone data. Compared to the existing review papers, this paper has a broader lens by reviewing the general applications of smartphone data before focusing on smartphone usage data. This paper also discusses some possible data sources that can be used in this research topic.

show abstract

Section: Smartphone Data Taxonomymentioning

confidence: 99%

Section: User Profilingmentioning

confidence: 99%

Section: User Profilingmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

A review on smartphone usage data for user identification and user profiling

Auliya

Nugroho

Setiawan

2021

CST

View full text Add to dashboard Cite

show abstract

“…Óskarsdóttir et al [18] demonstrate how including call networks information adds value in terms of profit by applying a profit measure and profit-based feature selection. Ots et al [19] also prove that mobile phone usage data can be used to make predictions and find the best classification method for credit scoring.…”

Section: B Mobile Phone Line Datamentioning

confidence: 99%

Feature-Level Fusion of Super-App and Telecommunication Alternative Data Sources for Credit Card Fraud Detection

Acevedo-Viloria¹,

Pérez²,

Jesus³

et al. 2021

Preprint

View full text Add to dashboard Cite

Identity theft is a major problem for credit lenders when there's not enough data to corroborate a customer's identity. Among super-apps-large digital platforms that encompass many different services-this problem is even more relevant; losing a client in one branch can often mean losing them in other services. In this paper, we review the effectiveness of a feature-level fusion of super-app customer information, mobile phone line data, and traditional credit risk variables for the early detection of identity theft credit card fraud. Through the proposed framework, we achieved better performance when using a model whose input is a fusion of alternative data and traditional credit bureau data, achieving a ROC AUC score of 0.81. We evaluate our approach over approximately 90,000 users from a credit lender's digital platform database. The evaluation was performed using not only traditional ML metrics but the financial costs as well.

show abstract

Algorithmic Credit Scoring in Vietnam: A Legal Proposal for Maximizing Benefits and Minimizing Risks

Laínez

Gardner

2023

AsianJLS

View full text Add to dashboard Cite

Artificial intelligence (AI) and big data are transforming the credit market around the world. Algorithmic credit scoring (ACS) is increasingly used to assess borrowers’ creditworthiness, using technology to glean non-traditional data from smartphones and analyze them through machine-learning algorithms. These processes promise efficiency, accuracy, and cost-effectiveness compared with traditional credit scoring. However, this technology raises public concerns about opacity, unfair discrimination, and threats to individual privacy and autonomy. Many countries in Southeast Asia are introducing ACS in consumer finance markets, although—even with the significant concerns raised—there is an ongoing and concerning lag in oversight and regulation of the process. Regulation is vital to delivering big data and AI promises in the financial services market, while ensuring fairness and public interest. This article utilizes Vietnam, where the lending industry deploys ACS but in a situation of legal limbo, as a case-study to analyze the consequences of this technology. Vietnam is one of the foremost Southeast Asian countries in which ACS usage is spreading rapidly, and this provides an excellent opportunity to review the regulation, or lack thereof, and determine the implications that this may have for other countries that are currently introducing ACS in consumer finance markets. The article concludes with a proposal to regulate ACS in Vietnam based on international regulation and guidelines on ACS, data privacy, and AI to enable a transparent, accessible, and fair process.

show abstract

Mobile Phone Usage Data for Credit Scoring

Cited by 8 publications

References 22 publications

A review on smartphone usage data for user identification and user profiling

A review on smartphone usage data for user identification and user profiling

Feature-Level Fusion of Super-App and Telecommunication Alternative Data Sources for Credit Card Fraud Detection

Algorithmic Credit Scoring in Vietnam: A Legal Proposal for Maximizing Benefits and Minimizing Risks

Contact Info

Product

Resources

About