Businesses gather a significant amount of valuable data on their prospects and customers. The challenge is that this data is often stored in multiple systems and can be difficult to gather and leverage.

Many people think of predictive modeling as just running an algorithm against data, but the reality is that the most complex work is in aggregating, analyzing, and presenting the data to the model so that the algorithm has the best information possible to work with. In order to be done well, this effort takes substantial knowledge and domain expertise with respect to both the data and the business problem.

Data from many diverse sources must be assembled in a manner conducive to modeling. Then, attributes are created that present the data to the modeling solution to predict outcomes. ChoicePoint Claims Solutions understands and has created the ChoicePoint ClaimFocus Solution Suite as a flexible suite of tools that can be customized to efficiently access different data sources. These data extraction tools are an important part of our predictive solution infrastructure.

ChoicePoint Claims Solutions utilize an advanced custom tool set to extract data from sources, dynamically merge the data and shape it into the format most suitable for a predictive application. Built upon our wide range of experience with various data applications, the tool set provides functionality to create workable data with execution efficiency, reduced resource consumption and minimal impact on a customer's operational system.

Our data tool set contains the flexibility to build customized data cleaning modules for normalization, transformation and formatting. This allows the predictive service to deal with various customers with different data. Our tool set can also automatically convert source data into a new data format allowing analysis application and predictive applications to work immediately.

Our tools are flexible enough to create new data features from source data based on business specific application requirements. Data features are a critical part in creating useful data marts, and our techniques allow us to shape original data with transformations and aggregations quickly into workable data.

Additionally, many of our ClaimFocus solutions utilize our powerful text analytics and linguistic extraction capabilities that allow the use of large amounts of information not available to non-text models. Our text based technology allows our solutions to understand and process free-form text, enabling our ability to leverage the vast amounts of information contained in nonstructured formats. This is where we uncover relevant information that is often embedded in narrative text and once identified is utilized within the model algorithm, enhancing the detection capability of a predictive modeling solution. Additionally, the names of recalled products such as water heaters, electrical appliances, or brands of automobile.

Our text analytics and linguistic extraction tools also allow for the daily scanning of text notes on the claim file to spot recovery opportunities and bring them to the attention of adjusters and subrogation professionals. This equates into fewer business rules being required and the need for fewer rules to maintain.