MARVEL AES is integral to CDM applications that search for information, compare data sets, and correct data errors based on reference libraries.


MARVEL AES


The MARVEL Application Engine Services (AES) provide two core services to CDM applications: intelligent search and data validation. Together, these services enable CDM applications to search for and compare multiple data sets, check operational data against reference data, and suggest corrections to invalid operational data. With these capabilities, users can quickly locate specific data and efficiently correct data errors.


MARVEL AES Intelligent Search can:


  • Use partial, imprecise, or even misspelled search terms to locate information.

    MARVEL AES can use partial, imprecise, or even misspelled search terms, as well as non-textual values (e.g., color), to locate information. It does so by mining data according to free-text elements (e.g., character sets and synonyms) and complex elements (e.g., geographic locations and dimensional quantities).


  • Mine data according to a similarity metric.

    MARVEL AES can search data according to similarity metrics that assess the degree of similarity between a user’s search criteria and the data being searched. This assessment is made possible by the MARVEL AES intelligent agents comparing underlying characteristics of a search against underlying characteristics of any number of data sets.


  • Weight search criteria.

    To increase the precision of its search, MARVEL AES enables users to rate the importance level of a search criterion value relative to another value. For example, if matching a color is more important than matching an ID number, the user can give twice as much weight to the color criterion over the ID number criterion in a search.


MARVEL AES Data Validation & Cleansing can:


  • Locate invalid data requiring corrections.

    MARVEL AES uses intelligent agents to check questionable operational data against accurate reference data values. If any discrepancies exist, MARVEL AES will immediately alert users, recommend corrections, and guide a user through the correction process. Once the corrections have been made, the user can then export the updated information to a wide range of file formats for external use.