IIT Database Group

Header bar


We conduct research that spans several areas of database systems such as information integration, data provenance, scheduling, stream processing, and data mining. Our main contributions impacted the theoretical and practical research in data provenance. We strive to develop solutions to emerging challenges in database systems such as tight integration of provenance support into database engines, provenance for distributed data processing paradigms, and provenance for sequences of database operations. For a list of our publications click here.


BART is an error-generation tool for data cleaning applications. Its purpose is to introduce errors into clean databases for the purpose of benchmarking data-repairing algorithms.
Explanations beyond Provenance
Explaining Query Results Beyond Provenance
Big Provenance
developing algorithms and systems for scaling provenance to Big Data dimensions.
A Hybrid OS / Runtime for Task-based In-memory Query Processing
A database-independent middleware for computing the provenance of queries, updates, and transactions
is a distributed database build from scratch that combines the best of traditional relational platforms with ideas from Big Data platforms.
A new, generic benchmark generator for data integration tasks.
LDV is a lightweight database virtualization system marrying OS and DB provenance.
PUGS is a unified framework for capturing why and why-not provenance of Datalog queries with negation and for automatic generation of concise provenance summaries.
Relevance-based Data Management
We use provenance to determine what data is relevant for which task and then exploit this information to improve a wide range of data management tasks.
Snapshot Semantics for Temporal Databases
We conduct a principled investigation of temporal query semantics. Our current focus is a provably correct approach for snapshot semantics for sets, bags, and beyond.
Uncertainty-Annotated Databases
In this project, we develop a practical, yet principled, approach for managing uncertain data.
Provenace for Updates and Transactions
In this project, we study provenance models for update and transactions and their implementation through reenactment, a declarative replay technique which utilizes audit logs and temporal database technologies.
A framework for user-friendly and effective data curation.

Past Projects

Computing fine-grained Provenance for Data Streams using Operator Instrumentation
Native Database Provenance
In this project we study how to integrate provenance techniques with a database core to improve various aspects of provenance managements including performance and storage requirements.
Declarative modelling and implementation of domain specific scheduling protocols.
Efficient Provenance Support for Relational Databases
Understanding the Behavior of Schema Mappings though Provenance and Meta-querying
Automatic generation of explanations for data exchange errors.


We would like to thank the following sponsors for their support:
Logo National Science Foundation

National Science Foundation

Logo Illinois Instituts of Technology

Illinois Instituts of Technology

Oracle External Research Office (ERO)


We are grateful to our awesome current and past collaborators!