Every day, millions of people use natural-language interfaces such as Siri, Google Now, Cortana, Alexa and others via in-home devices, phones, or messaging channels such as Messenger, Slack, and Skype, among others. The goal of this workshop is to advance the use of conversational natural language to query structured databases.
Building natural language interfaces for databases has been one of the "holy grails" of the database community since the early days of databases. In recent years, considerable progress has been made in this direction. Conversational interfaces are even more challenging since they need to keep track of context and deal with human short cuts.

Given the immense importance of such interfaces and the challenges that lie ahead in making such systems widely useful, building conversational interfaces over data has attracted interest from several areas: databases, machine learning, natural language processing and understanding and human-computer interaction forming a rich space of solutions.
The goal of this workshop is to bring together researchers and practitioners in this area, to clarify impactful research problems, share findings from large-scale real-world deployments, and generate new ideas for future lines of research. We seek papers that address any aspect of any of these issues.

Camera-Ready Instructions
Please use this modified vldb.cls when preparing your camera-ready file. Upload your final version until July 5, 2019.

Organizers and Chairs

  • H.V Jagadish, University of Michigan
  • Georgia Koutrika, Athena Research Center
  • Fatma Ozcan, IBM Research


Conversational Machines: Bridging the chasm between task-oriented and social conversations

Dilek Hakkani-Tur, Senior Principal Scientist at Amazon Alexa AI

Abstract: Conversational systems generally fall into two categories: task-oriented and social bots. Task-oriented systems aim to help users accomplish a specific task through multi-turn interactions, whereas socialbots focus on engaging and natural open-domain conversations. In natural interactions, even when conversation participants have a task or goal in mind, they can say things that are out of the boundaries of that task domain, and similarly they can switch to a task during a chitchat conversation. Hence, the ability to engage in knowledgeable social interactions and gracefully transition back and forth to the task is crucial for enabling natural conversations.

In this talk, I’ll summarize our recent work in both fronts, focusing on the convergence of approaches for the two categories of conversational systems. Starting with task-oriented interactions, I’ll present our approach for bootstrapping task-oriented dialogue systems from simulated seeker-provider conversations and dialogue state tracking using generate-and-copy mechanisms. This will be followed by a summary of learnings from previous Alexa Prize challenges and progression of our work as we approach the next challenge.

Short Bio: Dilek Hakkani-Tür is a senior principal scientist at Amazon Alexa AI focusing on enabling natural dialogues with machines. Prior to joining Amazon, she was leading the dialogue research group at Google, and worked as a principal researcher at Microsoft Research, International Computer Science Institute and AT&T Labs-Research. She received her BSc degree from Middle East Technical Univ. and MSc and PhD degrees in Computer Science from Bilkent University. Her research interests include conversational AI, natural language and speech processing, spoken dialogue systems, and machine learning for language processing. She has over 70 patents that were granted and co-authored more than 200 papers in natural language and speech processing. She is the Editor-in-Chief of the IEEE/ACM Transactions on Audio, Speech and Language Processing, and a fellow of the IEEE and ISCA.

Deploying a Conversational AI platform for Customer Support

Nikola Mrkšić, Co-founder and CEO, PolyAI

Abstract: PolyAI is a London-based startup with a leading machine learning platform for conversational agents. The deployed AI agents understand users, hold conversations without getting confused and can easily scale to new use cases or other languages. This talk will present the machine learning techniques that underpin the PolyAI platform and results from its early deployments in contact centre environments. [slides]

Short Bio: Nikola Mrkšić is the CEO and Co-Founder of PolyAI, a London-based Conversational AI company. Before starting PolyAI, Nikola worked with the Apple Siri team in Cambridge, and he was the first engineer at VocalIQ, a dialogue systems startup acquired by Apple. He did a PhD at Cambridge, working with Professor Steve Young at the Dialogue Systems Group.


  • - Understanding user intent and query semantics
  • - Explorations of ambiguity and other points of failure
  • - Contextualization (understanding the context of a query)
  • - Personalization (adapting to the individual posing the query)
  • - Ontologies, and other high-level data abstractions
  • - Entity matching
  • - intermediate representations
  • - Indexing techniques and other access methods
  • - Interoperability of conversational systems
  • - Portability of use history and personal profile
  • - Bootstrapping conversational workspaces
  • - Data driven conversation design
  • - Evaluation
  • - NL explanations
  • - Dialogue-based interaction, chatbots

Important Dates

Paper Submission: May 10 May 24, 2019 by 5pm PDT
Notification of Acceptance: June 15 June 20, 2019
Camera-ready Submission: July 5, 2019
Workshop Date: Friday, August 30, 2019

Submission Guidelines

Papers should be formatted according to the VLDB conference's camera-ready format, as embodied in the document templates. The standard paper length is 8 pages, including references and any appendices. If authors wish to preserve their ability to publish more complete versions of their work at other venues, they may choose to submit a 4-page paper instead. While this shorter paper will naturally have fewer details, it will be reviewed with the same quality standard as the standard-length paper, in terms of motivation, novelty and so on.

All accepted papers will have a presentation slot at the workshop. At least one author of every accepted paper is expected to attend the workshop to present the paper.

The conference management tool for the submission of abstracts and papers is accessible at: https://cmt3.research.microsoft.com/CAST2019

Accepted Papers

Title Author Names
Answering Complex Queries with Heterogeneous Structured Knowledge Sources extracted from Text Nikita Bhutani (University of Michigan, Ann Arbor, Michigan)*
Disambiguating Natural Language Queries with Tuples Christopher Baik (University of Michigan)*; Zhongjun Jin (University of Michigan); Michael Cafarella (University of Michigan)
Building a Hotel Concierge Bot: an industrial case study Behzad Golshan (Megagon Labs)*; George Mihaila (Megagon Labs); Chen Chen (Megagon Labs); Jonathan Engel (Megagon Labs); Alon Halevy (Megagon Labs); Yoshihiko Suhara (Megagon Labs); Wang-Chiew Tan (Megagon Labs); Michael Matuschek (TrustYou)
DBPal: Weak Supervision for Learning a Natural Language Interface to Databases Nathaniel Weir (Brown University); Andrew Crotty (Brown University)*; Alex Galakatos (Brown University); Amir Ilkhechi (Brown University); Shekar Ramaswamy (Brown University); Rohin Bhushan (Brown University); Ugur Cetintemel (Brown University); P. Ajie Utama (TU Darmstadt); Nadja Geisler (TU Darmstadt); Benjamin Hättasch (TU Darmstadt); Steffen Eger (TU Darmstadt); Carsten Binnig (TU Darmstadt)
Leveraging Human Learning in Interactive Data Exploration Sanad Saha (Oregon State University)*; Arash Termehchy (Oregon State University); Leilani Battle (University of Maryland)