Along with Big Data comes a unique set of business problems. From the obvious challenges of storing and accessing large data sets, to the bigger challenge of bringing together and interpreting disparate types of data. DataLabs effectively identifies and aligns solutions to address likely future enterprise needs.
Recent Experian DataLabs engagements
Proprietary recommender system that considers frequency, value, accuracy, diversity and novelty of data. One implementation runs more than 28 quintillion calculations and leverages insight from more than 5 billion transactions. The approaches include collaborative filtering and merchant similarity indexing.
Historical spend probability calculation and algorithms where individual profiles for consumers and small businesses are updated daily with new transactions; alerts are generated when events take place that are outside of expected behavior parameters. One implementation runs daily updates on more than 120 million individual profiles.
Income estimation derived from anonymized automated clearinghouse deposits using advanced clustering techniques. The process can identify more than 60 distinct income deposit patterns, and the output has been linked to assess direct deposit account liquidity risk and cross-sell product sequencing.
Other projects include evaluating the impact of multiple social media data sets on commercial credit risk, improving point proximity estimation using geolocation data and assessing consumer sensitivity to changes in product pricing.