Events Calendar

Iteratively Learning Data Transformation Programs from Examples

Speaker: Bo Wu, USC/ISI

Talk Title: Iteratively Learning Data Transformation Programs from Examples

Series: AI Seminar

Abstract: Data transformation is an essential preprocessing step in most data analysis applications. It often requires users to write many trivial and task-dependent programs. Recently, programming-by-example (PBE) approaches enable users to generate data transformation programs without coding. To correctly transform these datasets, existing PBE approaches typically require users to provide multiple examples to generate the correct transformation programs. These approaches time complexity grows exponentially with the number of examples and in a high polynomial degree with the length of the examples. Users have to wait a long time to see any response from the systems when they work on moderately complicated datasets. Moreover, existing PBE approaches also lack the support for users to verify the correctness of the transformed results.

To address the challenges, we propose an approach that generates programs iteratively, which exploits the fact that users often provide multiple examples iteratively to refine programs learned from previous iterations. We evaluated IPBE, the implementation of our iterative programming-by-example approach, against several state-of-the-art alternatives on various transformation scenarios. The results show that users of our approach used less time and achieved higher correctnesses compared to other alternative approaches.

Biography: Bo Wu is a newly graduated PhD from University of Southern California. He worked at Information Integration group at Information Science Institute. His research focuses on automatically generating data transformation programs. He received his B.S. in software engineering from Harbin Institute of Technology and his M.S. in computer science from Institute of Computing Technology, Chinese Academy of Sciences.

Host: Craig Knoblock

Location: Information Science Institute (ISI) - 1135 - 11th fl Large CR

Audiences: Everyone Is Invited

Contact: Alma Nava / Information Sciences Institute