Unlike traditional data obtained from insurance providers, criminal records possess complexities that make data acquisition and delivery very challenging. With more than 3,000 counties in the United States, even more municipal jurisdictions, departments of correction, federal courts, and myriad other sources, it is a time consuming and sophisticated process to amass the data necessary to establish a national criminal database.

The end-to-end process may sound easy, but mastering the art of data compilation for client reporting requires the effort of multiple resources and areas of expertise. Add to this a political environment in which laws change frequently, court clerks who control data access, and severely restricted governmental funding to build the technology to provide data in a consistent, clear, and concise format. In this light, the data-compilation process becomes even more complex. But it’s not impossible.

Since no single source exists to acquire criminal-history information, a variety of data-gathering options are available to a criminal-data aggregator. Bulk data is obtained through a process of purchasing and receiving information from a source that compiles large volumes of criminal records. Another technique is to retrieve data online directly from a source. And where online or bulk data is not available, the option exists to conduct an on-site search directly at a courthouse by utilizing the services of a “court runner” to retrieve actual records manually from archives.

Criminal-records data received from a source is generally inconsistent and unstructured. This data is not used in transaction processing systems—but is a historical record of criminal proceedings. Therefore, a custom interface to manage data integration, quality control, and data monitoring and reporting is advantageous. A considerable amount of the data can be entered as free-form text with very few discrete data elements. Due to the nature of criminal records, only certain fields—needed to provide matches on search criteria—can be extracted. The remaining data is stored as text blobs in the database. Before each source file is processed, a quality-assurance check is performed on the formatting and record counts, which provides insight into any changes that regularly occur with data files. Each file can then be processed using a unique data map stored within an ETL (extract, transform, and load) software package.

After the data has been loaded, a standard set of queries can be used for a quality-assurance check. Special “fill reports” are then prepared that describe the data. The reports provide statistics on data quality at the state and county level. This ensures all data was loaded properly—and any newer data supersedes previous data with respect to quality in the appropriate fields.

The vast majority of criminal records do not contain a single identifier like a Social Security number. Therefore, searching for records using first, middle, and last names is critical to the process deployed by providers of background-check data. Since names are often mistyped or abbreviations and nicknames are commonly substituted for proper names, searching on “name only” presents a challenge. To help with this, leading providers employ sophisticated “fuzzy name” technology in their processes. This unique technology, available to providers such as IntelliCorp, significantly enhances the solution’s ability to find matching records related to a query.

Leading-edge providers of background checks build multiple products and services based on proprietary criminal databases. They are provided to clients through a variety of secure channels. A typical criminal background check starts with a business or organization obtaining a signed release from the applicant, then providing the subject’s information to conduct the search. This typically includes name, Social Security number, date of birth, and address. Preemployment screening is a process regulated by the Fair Credit Reporting Act (FCRA), and providers must follow the federal regulations when completing a background check. The technical processes include flagging and tracking information in the proprietary database, as well as real-time filtering of information that cannot be used in the hiring process. The criminal background check is then processed by extracting any information matching the subject search criteria and constructing a report of this information for the client. The report is then delivered to the client securely, generally via a website, web service gateway, or encrypted e-mail.

Since the company’s inception, IntelliCorp has aggregated criminal data received from court systems and other reporting agencies to build a unified database. It then provides the information to businesses and organizations to mitigate risk. Of all IntelliCorp offerings, criminal records are the highest-volume subset of products ordered by clients. William F.Hauswirth, CPCU, ARM, is president of Intellicorp Records, Inc., an ISO company based in Beachwood, Ohio. IntelliCorp (www.intellicorp.net) helps businesses and organizations reduce exposure to fraud, workplace violence, and negligent-hiring lawsuits.