Experian Information Solutions, Inc.

Embodiments are disclosed of a system for detecting potential online transaction fraud. The system can receive transaction data associated with an online transaction. In some embodiments, the system can determine consumer identification data and payment instrument data from the transaction data. The system can compare the consumer identification data and the payment instrument data associated with the online transaction against consumer profiles generated based on existing consumer information and payment instrument information. Based on the comparison, the system can generate a match score indicative of a likelihood that the online transaction is fraudulent.

Methods and systems are provided for generating alerts in association with a request for credit data or an update to credit data. For example, information identifying an individual may be received from a credit bureau when an individual applies for a product or service from a requesting entity. The requesting entity may be a product or service provider that requests credit data of potential purchasers of its products. An electronic notification may then be sent in association with an event object generated by a notification system that is broadcast for delivery to one or more other systems, where the event object may be broadcast substantially in real time, such as via an enterprise server bus, relative to the requesting entity's initial request.

The systems and methods described herein allow consumers to lock or unlock their credit files at multiple credit bureaus in real-time or near real-time. The service may allow a consumer to provide identifying information, such as a personal identifier to lock or unlock credit files at a plurality of credit bureaus over a network. Upon receiving the personal identifier, the system may use the personal identifier to translate the identifier into a plurality of access codes for respective credit bureaus, for example by accessing a data structure, such as a database or table, which stores a personal identifier and access codes that are associated with a consumer. The system may then use the access codes to automatically initiate locking or unlocking of credit files for the consumer at the respective credit bureaus.

One embodiment of the system and method described herein provides a score generator system that generates an automated vehicle specific valuation of a used car based on the physical and historical attributes of that vehicle. This score may indicate the likelihood that the vehicle will be on the road in a specific period of time. The score may give an absolute percentage of such likelihood or it may give a value relative to all other used vehicles in a database, all other used vehicles of the same make/model/year, or a certain subset of the vehicles in a database. In one embodiment, the score generator system includes a data link module for linking vehicle data and filter module for applying a multi-level filters that process the linked vehicle data.

A data processing system is disclosed for accessing databases and updated data items and triggering event notifications. The data processing system may comprise a first database including a plurality of records, and a second database including a plurality of trigger indicators. The database system may further include a hardware processor configured to execute computer-executable instructions in order to: receive an update data item; identify a record corresponding to the update data item; cause an update to the record based on information included with the update data item; identify a trigger indicator corresponding to the update to the record; determine that a type of the trigger indicator matches a type of the update to the record; and generate an event notification including information included in the update.

Systems and methods are described for machine learning-based generation of scorecards and boosted decision trees that facilitate explainable predictions. A scorecard machine learning model may be applied to historical records such that the model, for each of a number of variables, automatically generates (a) normal bins for normal values of the variable that fall within a valid range of values and (b) at least one special bin for special values of the variable that fall outside the valid range of values. Adjacent bins of the normal bins may be separated by a threshold value and each normal bin and each special bin may have an assigned score value. A risk assessment score may be generated based at least in part on the model identifying the assigned value score for each of the variables based on the normal or special bin to which each variable is assigned.

A digital targeting system is configured to access data to determine users that satisfy content factors, such as pre-validation criteria, based on user information stored in a database having restrictions on sharing of user data, such as government regulations on sharing of the user information. The digital targeting system may then generate a pre-validated user ID list indicating those users that match the content factors. Users on the pre-validated user ID list may then be matched to users in other databases, such as a digital ID database, to identify digital IDs for the users that are usable by a digital display entity.

This disclosure relates to data association, attribution, annotation, and interpretation systems and related methods of efficiently organizing heterogeneous data at a massive scale. Incoming data is received and extracted for identifying information (“information”). Multiple dimensionality reducing functions are applied to the information, and based on the function results, the information are grouped into sets of similar information. Filtering rules are applied to the sets to exclude non-matching information in the sets. The sets are then merged into groups of information based on whether the sets contain at least one common information. A common link may be associated with information in a group. If the incoming data includes the identifying information associated with to the common link, the incoming data is assigned the common link. In some embodiments, incoming data are not altered but assigned into domains.

Systems and methods are provided for securely replicating data, such as from an on-premises database to remote network-accessible storage, with high throughput and robust security mechanisms. Messages or events may be published via a first message queue service, such as an on-premises message queue service, that are then encrypted, optionally at the field level, before being passed to a managed message queue service. The encrypted messages may then be processed by a replication application configured to split batches of messages into sub-batches that are processed in parallel before being committed to a managed database at the batch level as individual batches complete processing.

Embodiments of a system and method are described for generating a finance attribute. In one embodiment, the systems and methods retrieve raw tradeline data from a plurality of credit bureaus, retrieve industry code data related to each of the plurality of credit bureaus, determine one or more tradeline leveling characteristics that meet at least one pre-determined threshold, and generate a finance attribute using the selected leveling characteristics.