Credit Risk Evaluation Based on Social Media

Fei, Wenying; Gu, Jing; Yang, Yang; Zhou, Zongtan

doi:10.1016/j.procs.2015.07.165

Cited by 14 publications

(6 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Another approach, based on social media, and proposed by Yang and Zhou, is shown in [3], which analyzes whether the different opinions shared by users about different financial topics are relevant or not for the prediction of credit risk. Opinions were analyzed from two social networks for financial investors in China, in addition to various articles published by financial analysts.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

The Naïve Associative Classifier With Epsilon Disambiguation

et al. 2020

View full text Add to dashboard Cite

This paper presents the Naïve Associative Classifier with Epsilon disambiguation (NACε), an extension of the Naïve Associative Classifier that, by including a procedure to disambiguate classes in regions where Bayes risk is high, has a positive effect on the performance of the classifiers of the associative approach on several datasets belonging to the financial environment, particularly in terms of credit risk. The experiments conducted to test the NACε were based on 12 datasets composed with financial information and associated with five stages of the credit process: promotion, evaluation, granting, monitoring and recovery. Due to the severe imbalance present in most datasets, the performance of the proposed algorithm was measured using the area under the ROC curve. Likewise, 5 × 2 stratified cross validation was made and finally a couple of statistical tests were applied to compare the results. After applying the NACε to the datasets, a successful disambiguation of classes was observed. In the real world this fact could help financial institutions to evaluate the credit applications more effectively and thus, contribute to the mitigation of monetary losses derived from the poor quality of the information.

show abstract

Section: Related Workmentioning

confidence: 99%

“…Intelligent Computing (IC) algorithms are useful when searching computational solutions to financial risk related problems. Currently, several research papers address this type of problem; specifically: credit risk [2], [3], bankruptcy [1], [4], [5] or marketing [6], [7], among others.…”

Section: Introductionmentioning

confidence: 99%

The Naïve Associative Classifier With Epsilon Disambiguation

et al. 2020

View full text Add to dashboard Cite

show abstract

“…Social media will influence individual investment decisions. Yang Y, Gu J et al [38] chose the two most representative social platforms in Chinese financial area and selected behavioural data to do textual analysis, which proved that the results of credit risk prediction using behavioural data were more accurate than those predicted by professional risk managers. Behavioural data is applied not only in personal credit scoring, but also in other research areas.…”

Section: Literature Reviewmentioning

confidence: 99%

“…There are some cheats or people who spread false information on the social platform, so we should find the right way to judge users with good credit and users with bad credit [1,11]. By analyzing the huge behavioural data on the social platform, we can achieve diverse purposes including evaluating personal credit of users in Social Network Sites (SNS) [12].…”

Section: Introductionmentioning

confidence: 99%

Credit Scoring with AHP and Fuzzy Comprehensive Evaluation Based on Behavioural Data from Weibo Platform

Yao

Chen

2019

Teh. vjesn.

View full text Add to dashboard Cite

It is increasingly necessary to evaluate the customers' credit. In the era of big data, Information on the Internet is commonly used to judge the credit worthiness of customers. Some users' credit information is incomplete or unavailable, so credit managers cannot judge the true credit situation of these users. However, with the support of social data especially behavioural data and credit evaluation system, this problem can be effectively solved. This study used Weibo to obtain the behavioural data of Chinese users for credit evaluation. Two methods are used to calculate the credit scores of Weibo users, which are the analytic hierarchy process (AHP) and fuzzy comprehensive evaluation methods. By analysing social processes and inviting experts to make decisions, we constructed a credit evaluation system to expose users' behavioural characteristics. We found that the three key indexes determining the user's social credit are personal identification, behavioural characteristics and interaction among friends. Then, AHP was used to determine the weight of each index. Finally, a static algorithm was proposed to compute the credit evaluation system of Weibo users using fuzzy comprehensive evaluation methods.

show abstract

“…Even in today's standards, the traditional approach, which uses only hard information, is that which is widely used by firms but there is lack of studies that analyze textual information (Fei et al, 2015 ; Allahyari et al, 2017 ). Jiang et al ( 2018 ) demonstrate how the use of textual data can increase the predictive power of a model, combining soft information with typically financial information analyzing the main p2p platforms in China.…”

Section: Introductionmentioning

confidence: 99%

On the Improvement of Default Forecast Through Textual Analysis

Cerchiello

Scaramozzino

2020

Front. Artif. Intell.

View full text Add to dashboard Cite

Textual analysis is a widely used methodology in several research areas. In this paper we apply textual analysis to augment the conventional set of account defaults drivers with new text based variables. Through the employment of ad hoc dictionaries and distance measures we are able to classify each account transaction into qualitative macro-categories. The aim is to classify bank account users into different client profiles and verify whether they can act as effective predictors of default through supervised classification models.

show abstract

Credit Risk Evaluation Based on Social Media

Cited by 14 publications

References 13 publications

The Naïve Associative Classifier With Epsilon Disambiguation

The Naïve Associative Classifier With Epsilon Disambiguation

Credit Scoring with AHP and Fuzzy Comprehensive Evaluation Based on Behavioural Data from Weibo Platform

On the Improvement of Default Forecast Through Textual Analysis

Contact Info

Product

Resources

About