GGrantIndex
← Search

CAREER: Reverse Data Management: Reverse-Engineering Data Transformations to Understand, Diagnose, and Manipulate Data

$576,092FY2015CSENSF

University Of Massachusetts Amherst, Amherst MA

Investigators

Abstract

Today, data is critical in almost every aspect of society, including healthcare, education, economy, and science. Applications in these domains have benefited from the rapid growth and increased availability of data through online repositories, marketplaces, and other sharing platforms. However, because data is easily shared and reused, it has become less curated and less reliable. Data is often misused because its validity and origin are unclear, and mistakes easily propagate as data is often used to derive other data. This project will design and develop new technologies to assist data-reliant applications in overcoming these challenges, by extending data management capabilities to support operations that reverse-engineer data transformations. Understanding and reasoning with data derivations will lead to better understanding of surprising observations, more accurate and efficient diagnoses of problems, and better tools for making data-driven decisions and planning. This work will broadly impact society through the development of new data analysis capabilities, which will, in turn, improve data quality and positively affect data-driven processes. Overall, these improvements will increase the positive impact that data and computing systems have on society. This project will establish the theoretical foundations and will address practical challenges to support reverse data management capabilities in database systems. Reverse data management focuses on reverse transformations that manipulate input data on behalf of desired outcomes in the output data (as opposed to traditional data management that typically focuses on forward-moving data flows, such as querying, aggregation, and indexing). The research goal of this project is to augment database systems with user-facing reverse data management functionality. This project will target three reverse data management needs of modern applications: (1) assisting users with understanding data and verifying data-based computation results, (2) diagnosing problems by tracing the root causes of data errors, and (3) identifying changes in data and processes that lead to improved, more desirable results. For further information see the project web site at: http://rdm.cs.umass.edu

View original record on NSF Award Search →