PossiblePasts

The ThinkStraight start-up had identified an interesting data forensic problem in the specialized context of financial (accounting) data. The project allowed us to analyze the scientific questions raised by their problem, which we found novel, and draw a state of the art of the two closest areas, namely data cleaning and data quality analysis. We have also analyzed the software architecture that ThinkStraight was planning to build and compared it to existing technologies developed within the team, notably the LN2R and EdiFlow prototypes; we found that the ThinkStraight problem required quite radically different functionalities and given the very early stages of the company’s development, the collaboration was not further pursued.