RAT-SQL: Relation-Aware Schema Encoding and Linking for Text-to-SQL Parsers
- Bailin Wang, Richard Shin, Xiaodong Liu, Oleksandr Polozov, Matthew Richardson
- Computer ScienceAnnual Meeting of the Association for…
- 10 November 2019
This work presents a unified framework, based on the relation-aware self-attention mechanism, to address schema encoding, schema linking, and feature representation within a text-to-SQL encoder and achieves the new state-of-the-art performance on the Spider leaderboard.
FlashMeta: a framework for inductive program synthesis
- Oleksandr Polozov, Sumit Gulwani
- Computer ScienceConference on Object-Oriented Programming Systems…
- 23 October 2015
The FlashMeta framework implements a novel program synthesis methodology, allowing synthesizer developers to generate an efficient synthesizer from the mere DSL definition (if properties of the DSL operators have been modeled), and found that 10+ existing industrial-quality mass-market applications based on PBE can be cast as instances of D4.
Learning Syntactic Program Transformations from Examples
- Reudismam Rolim de Sousa, Gustavo Soares, B. Hartmann
- Computer ScienceInternational Conference on Software Engineering
- 31 August 2016
REFAZER builds on the observation that code edits performed by developers can be used as input-output examples for learning program transformations, and leverages state-of-the-art programming-by-example methodology to learn transformations that can fix other students' submissions with similar faults.
SCoRe: Pre-Training for Context Representation in Conversational Semantic Parsing
- Tao Yu, Rui Zhang, Oleksandr Polozov, Christopher Meek, A. Awadallah
- Computer ScienceInternational Conference on Learning…
- 2021
Generative Code Modeling with Graphs
- Marc Brockschmidt, Miltiadis Allamanis, Alexander L. Gaunt, Oleksandr Polozov
- Computer ScienceInternational Conference on Learning…
- 22 May 2018
A novel model is presented that uses a graph to represent the intermediate state of the generated output and can generate semantically meaningful expressions, outperforming a range of strong baselines.
Neural-Guided Deductive Search for Real-Time Program Synthesis from Examples
- Ashwin J. Vijayakumar, Abhishek Mohta, Oleksandr Polozov, Dhruv Batra, Prateek Jain, Sumit Gulwani
- Computer ScienceInternational Conference on Learning…
- 15 February 2018
This work proposes Neural Guided Deductive Search (NGDS), a hybrid synthesis technique that combines the best of both symbolic logic techniques and statistical models and produces programs that satisfy the provided specifications by construction and generalize well on unseen examples, similar to data-driven systems.
Structure-Grounded Pretraining for Text-to-SQL
- Xiang Deng, Ahmed Hassan Awadallah, Christopher Meek, Oleksandr Polozov, Huan Sun, Matthew Richardson
- Computer ScienceNorth American Chapter of the Association for…
- 24 October 2020
A novel weakly supervised Structure-Grounded pretraining framework for text-to-SQL that can effectively learn to capture text-table alignment based on a parallel text- table corpus and brings significant improvement over BERTLARGE in all settings.
User Interaction Models for Disambiguation in Programming by Example
- M. Mayer, Gustavo Soares, Sumit Gulwani
- Computer ScienceACM Symposium on User Interface Software and…
- 5 November 2015
This work presents two novel user interaction models that communicate actionable information to the user to help resolve ambiguity in the examples of PBE systems.
Program Synthesis and Semantic Parsing with Learned Code Idioms
- Richard Shin, Miltiadis Allamanis, Marc Brockschmidt, Oleksandr Polozov
- Computer ScienceNeural Information Processing Systems
- 26 June 2019
PATOIS is a system that allows a neural program synthesizer to explicitly interleave high-level and low-level reasoning at every generation step, and it accomplishes this by automatically mining common code idioms from a given corpus and incorporating them into the underlying language for neural synthesis.
Execution-Guided Neural Program Decoding
- Chenglong Wang, Po-Sen Huang, Oleksandr Polozov, Marc Brockschmidt, Rishabh Singh
- Computer ScienceArXiv
- 9 July 2018
A neural semantic parser that translates natural language questions into executable SQL queries with two key ideas, including an encoder-decoder model, and using the execution semantics of SQL to repair decoded programs that result in runtime error or return empty result.
...
...