Learning a Neural Semantic Parser from User Feedback

@inproceedings{Iyer2017LearningAN,
  title={Learning a Neural Semantic Parser from User Feedback},
  author={Srini Iyer and Ioannis Konstas and Alvin Cheung and Jayant Krishnamurthy and Luke Zettlemoyer},
  booktitle={ACL},
  year={2017}
}
Highlight Information
We present an approach to rapidly and easily build natural language interfaces to databases for new domains, whose performance improves over time based on user feedback, and requires minimal intervention. [...] Key Method These models are immediately deployed online to solicit feedback from real users to flag incorrect queries. Finally, the popularity of SQL facilitates gathering annotations for incorrect predictions using the crowd, which is directly used to improve our models.Expand Abstract

Citations

Publications citing this paper.
SHOWING 1-10 OF 108 CITATIONS

DBPal: Weak Supervision for Learning a Natural Language Interface to Databases

VIEW 11 EXCERPTS
CITES METHODS & BACKGROUND
HIGHLY INFLUENCED

Learning natural language interfaces with neural models

VIEW 9 EXCERPTS
CITES BACKGROUND & METHODS
HIGHLY INFLUENCED

An End-to-end Neural Natural Language Interface for Databases

VIEW 10 EXCERPTS
CITES BACKGROUND & METHODS
HIGHLY INFLUENCED

Improving Text-to-SQL Evaluation Methodology

VIEW 9 EXCERPTS
CITES METHODS & BACKGROUND
HIGHLY INFLUENCED

Robust Text-to-SQL Generation with Execution-Guided Decoding

VIEW 8 EXCERPTS
CITES METHODS & BACKGROUND
HIGHLY INFLUENCED

Don't paraphrase, detect! Rapid and Effective Data Collection for Semantic Parsing

VIEW 5 EXCERPTS
CITES BACKGROUND & METHODS
HIGHLY INFLUENCED

Explaining Natural Language query results

VIEW 14 EXCERPTS
CITES BACKGROUND
HIGHLY INFLUENCED

Lifecycle of neural semantic parsing

VIEW 6 EXCERPTS
CITES BACKGROUND & METHODS
HIGHLY INFLUENCED

One-Shot Learning for Text-to-SQL Generation

VIEW 10 EXCERPTS
CITES METHODS, BACKGROUND & RESULTS
HIGHLY INFLUENCED

FILTER CITATIONS BY YEAR

2016
2020

CITATION STATISTICS

  • 14 Highly Influenced Citations

  • Averaged 34 Citations per year from 2017 through 2019

  • 23% Increase in citations per year in 2019 over 2018

References

Publications referenced by this paper.
SHOWING 1-10 OF 37 REFERENCES

Building a Semantic Parser Overnight

VIEW 4 EXCERPTS
HIGHLY INFLUENTIAL

PPDB: The Paraphrase Database

VIEW 4 EXCERPTS
HIGHLY INFLUENTIAL

Learning Dependency-Based Compositional Semantics

VIEW 4 EXCERPTS
HIGHLY INFLUENTIAL

Adam: A Method for Stochastic Optimization

VIEW 2 EXCERPTS
HIGHLY INFLUENTIAL

Grounded Unsupervised Semantic Parsing

VIEW 3 EXCERPTS
HIGHLY INFLUENTIAL