Advances in technology and software are enabling organizations to adopt more powerful and advanced analytical approaches to help drive growth, efficiency, competitiveness and manage risk. The availability of new data sources creates enormous potential for developing predictive credit risk management models which go far beyond ‘traditional’ internal and credit bureau data.
Experian has already invested in this area, developing extensive knowledge, experience and established a track record in delivering big data, data science, advanced analytics and smart credit solutions, in numerous countries across the globe.
We’ve briefly outlined an advanced analytics solution- Web Data Insights(WDI) - that uses public unstructured data gathered from the web, combined with machine learning techniques, to develop models that enable companies to add another dimension to the credit risk matrix for portfolio assessment in the whole credit lifecycle from origination to customer insights, collections and fraud prevention.
Improved risk assessment with the addition of public non-traditional and unstructured data
Ability to highlight root causes of high-risk events
Accurate prediction about future events and a customer’s likely propensity to buy, churn, or engage
Reduction in bad debt, improvements in acceptance rates and in the quality of accepted applications
Early warning signals and deeper portfolio insights
Improved statistical models based on machine learning techniques
Our comprehensive end-to-end WDI solution covers web crawling, site classification, text mining, segmentation analysis and machine learning models. Each account from a pre-defined sample is analysed, searched for on the internet with publicly-available information collected – an automated process known as ‘crawling’ or web datascraping.
Information is then classified under various categories by sophisticated text mining algorithms, before meaningful and relevant information is converted into usable data. Differing text mining and sentimental analysis methods can be applied to order unstructured data. From there, predictive variables base.
"Web Data Insights assess customer behavior through on-line presence measurement. It enables companies to add value to their existing matrix for portfolio evaluation.
Web Data Insight: Credit Risk Management for SMEs portfolio
Accessing advanced analytics for better insights
Combine non-traditional data sources with machine learning to create long-lasting customer relationships