Generation of realistic mock data

#Software providers commonly develop software applications that deal with large amounts of data stored in databases, such as relational databases, graph databases, object-oriented databases, etc. While the software applications may have been rigorously verified and validated by the software providers using standard or proprietary test data, such test data may not be representative of realistic data collected by users of the software applications. Even if the test data is representative of the realistic data, users of the software application may handle the data in non-standard manners that deviate from standard data-handling procedures that have been tested during software development. As a result, users of the software applications may encounter technical problems when using the software applications to manage or process realistic data that are specific to the users' domain. To help diagnose and/or resolve the technical problems experienced by the users, software providers often need to replicate such problems. However, it may not be feasible for the software providers to access the user's data for a variety of reasons, e.g., the volume of the data is too large, the user's data are confidential, the user's data are protected by statutory privacy regulations, and so on. Thus, there remains a need for intelligent generation of realistic mock data that can accurately simulate the realistic data encountered by the users.

#This Summary is provided to introduce a selection of concepts in a simplified form that are further described below in the Detailed Description. This Summary is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to be used to limit the scope of the claimed subject matter.

#Certain embodiments of the disclosure concern a computer-implemented method. The method can include receiving statistical artefacts representing an original database table which has original values arranged in a plurality of rows and a plurality of columns, and instantiating a reconstructed database table based on the statistical artefacts. The reconstructed database table can include reconstructed values arranged in at least a subset of the plurality of columns of the original database table. The statistical artefacts can include first metrics characterizing distributions of row-based sequences of the original values and second metrics characterizing distributions of the first metrics. Distributions of row-based sequences of the reconstructed values in the reconstructed database table can be consistent with the first metrics, and meta-distributions of row-based sequences of the reconstructed values in the reconstructed database table can be consistent with the second metrics.

#FIG. 1 is an overall block diagram of an example system for reconstructing database tables in a reconstructor computing system that preserve statistical properties of original database tables in a source computing system.

#FIG. 2 is a flowchart illustrating an example overall method of reconstructing database tables in a reconstructor computing system that preserve statistical properties of original database tables in a source computing system.

#FIG. 3 is an example embodiment comparing an original database table and a reconstructed database table.

#Enterprise software generally maintains a database environment that can include a plurality of database tables. Users of enterprise software can run one or more queries against the database tables. Such queries can retrieve objects from the database tables that satisfy one or more conditions predefined in the queries.

#Enterprise resource planning (ERP) systems allow an organization to use integrated software applications to manage and automate many back-office functions related to technology, services and human resources. ERP software typically integrates day-to-day processes of an enterprise and core capabilities in a single database, application and user interface.

#Users of ERP software may encounter technical problems that require customer support from customer support provider (e.g., ERP software provider or other third party) for debugging and resolution. As is often the case in software, the root cause of such problems can sometimes be difficult to detect.