Research

A conversational data system (or intelligent data assistant) understands user queries in natural language and uses natural language to explain queries, ask clarifications from the user, and describe results. Imagine, an intelligent data assistant, that allows us to express any query on any data in natural language, to dig into data and find answers by conversing and collaborating with the system, just like working with a human. Such systems are much more complex than any “digital assistant” we know of today, which are based on pre-specified dialogues and are able to answer simple questions like “what are the open pharmacies in my neighborhood?”. They require the synergy of several technologies and innovation in all these fronts, including natural language interfaces, data exploration, and conversational AI.

We are envisioning tools that enable a guided, interactive dialogue with the user to help users quickly discover data parts or insights of interest. We consider this process bound by a two-way communication where: (a) an intelligent system discovers and recommends interesting data/insights, tailored to the user needs, and (b) the user interacts with the system, providing feedback that guides the exploration process. Our goal is to study the challenges that arise from this exploration paradigm, develop new interactive data exploration techniques that combine efficiency with effectiveness, develop methodologies for the evaluation of such systems, and systematically evaluate our algorithms and systems using different data and use cases.

Our goal is to leverage the best of both worlds, data management and deep learning, to build systems that can learn from user queries and from data to not only process queries more efficiently but also to understand user intention, adapt to users, and help the user achieve their goal more effectively.