Human Data Interaction Lab

What are we about?

The goal of the Human-Data Interaction Lab (HuDa) is to research methods at the intersection of data management, data science, and interactive systems to enable the derivation of insights from data and to support data-driven decision-making. Our approach is human-centric in that we aim to understand how humans at any level of data-literacy (which include analysts, scientists, journalists, researchers, data enthusiasts) interact with data. Through this understanding, we build formalizations and systems that that make it easier for humans to effectively interact with data. HuDa, thus, addresses the critical bottleneck in today’s data analysis: this bottleneck is not the lack of data, or our ability to analyze it at scale, but the lack of human cognition and time to build and understand data processes.

The HuDa Interaction Lab has made several research contributions in the following areas in the past years

• Novel querying paradigms for prescriptive analytics.
• Example-driven Interfaces for data tasks.
• Debugging complex data analytics.
• Teaching and democratizing data science.

Incidentally, Huda in arabic means guidance. We hope to build tools that guide humans through better data analysis and data-driven decision-making.

Active Research Areas

Past Projects

Data-Driven Decision Making in Health

Suggesting the best next course of action for patients in ICUs

Dennis Shasha, Azza Abouzied, Hashim Hayat, Rick Hull, Elisa Sorrentino, Farah Shamout, Abbas Shojaee, Yifan Li

EpiPolicy

Policy Optimization to Control Epidemic Outbreaks

Miro Mannino, Anh Mai, Zain Tariq, Dennis Shasha, Azza Abouzied

Hill and Dale

Streaming queries with AQuery

Ahmad Chatha, Kostas Zoumbatianos, Dennis Shasha, Azza Abouzied

Impact3D

Visualizing the impact of Charitable Donations with Augmented Reality

Lauren McMillen, Azza Abouzied

Qetch

Expressive Time Series Querying with Hand-Drawn Scale-Free Sketches

Miro Mannino, Azza Abouzied

SEER

Auto-Generating Information Extraction Rules from User-Specified Examples.

Maeda Hanafi, Azza Abouzied, Laura Chiticariu, Yunyao Li

Synner

Generating Realistic Synthetic Data

Miro Mannino, Azza Abouzied

Teaching Query Languages

Using Visual Query Plans and Data to Explain Relational and Graph Queries

Yousra Hassan, Miro Mannino, Azza Abouzied, Semih Salihoglu

Texture

A Collaborative Framework for Structure Identification over Print Documents

Maeda Hanafi, Miro Mannino, Azza Abouzied

Validating Generative Models

Evaluating the realism of synthetic data sets generated by GANs and language models

Miro Mannino, Ofer Mendelevitch, Azza Abouzied

Researchers

Past Students & Researchers

Teaching

In our efforts to increase data-literacy and engage students in human-data interaction research, we have designed and taught the following courses:

• An Online Database Systems Course for NYUAD
• Core X Data
• Capstone: Systems for Interactive Data Analysis

Join Us

Want to a be Research Assistant, post-doc, graduate or undergraduate researcher at the HuDa Interaction Lab - we are hiring! Please email the PI.