Panupong Pasupat's Personal Page

Advisor: Percy Liang

Areas of Interest: Natural Language Processing

Office: Gates Building, Room 256

Resources

SEMPRE: Our framework for training semantic parsers from denotations.
WikiTableQuestions Dataset: A dataset of complex question on Wikipedia tables.
- SEMPRE models trained on the dataset.
- DPD Outputs: Inferred consistent logical forms for the training examples in the WikiTableQuestions dataset.
SCONE: Sequential CONtext-dependent Execution dataset.
MiniWoB++ Benchmark: A benchmark for learning reinforcement learning agents that interact with websites. An extension of OpenAI MiniWoB benchmark.

Mapping Natural Language Commands to Web Elements
Panupong Pasupat, Tian-Shun Jiang, Evan Liu, Kelvin Guu, Percy Liang
Empirical Methods on Natural Language Processing (EMNLP), 2018
paper dataset + code CodaLab
Reinforcement Learning on Web Interfaces using Workflow-Guided Exploration
Evan Zheran Liu*, Kelvin Guu*, Panupong Pasupat*, Tianlin Shi, Percy Liang (*equal contributions)
International Conference on Learning Representations (ICLR), 2018
paper code CodaLab dataset
Zero-Shot Entity Extraction from Web Pages
Panupong Pasupat, Percy Liang
Association for Computational Linguistics (ACL), 2014
paper project + dataset slides

Macro Grammars and Holistic Triggering for Efficient Semantic Parsing
Yuchen Zhang, Panupong Pasupat, Percy Liang
Empirical Methods on Natural Language Processing (EMNLP), 2017
paper CodaLab slides
Errata: In Equation 9, the term in the logarithm should be p_i⁺ / p_i^- instead of p_i⁺ / (p_i⁺ + p_i^-).
Version 2 on arXiv has the corrected formula.
From Language to Programs: Bridging Reinforcement Learning and Maximum Marginal Likelihood
Kelvin Guu, Panupong Pasupat, Evan Zheran Liu, Percy Liang
Association for Computational Linguistics (ACL), 2017
paper dataset CodaLab | More on Kelvin's website
Inferring Logical Forms From Denotations
Panupong Pasupat, Percy Liang
Association for Computational Linguistics (ACL), 2016
paper CodaLab slides computed output
Simpler Context-Dependent Logical Forms via Model Projections
Reginald Long, Panupong Pasupat, Percy Liang
Association for Computational Linguistics (ACL), 2016
paper dataset CodaLab
Compositional Semantic Parsing on Semi-Structured Tables
Panupong Pasupat, Percy Liang
Association for Computational Linguistics (ACL), 2015
paper project + dataset CodaLab slides poster blog post

Unsupervised Relation Detection Using Automatic Alignment of Query Patterns Extracted from Knowledge Graphs and Query Click Logs
Panupong Pasupat, Dilek Hakkani-Tür
Interspeech, 2015
paper slides extended work
Query Understanding Enhanced by Hierarchical Parsing Structures
Jingjing Liu, Panupong Pasupat, Yining Wang, Scott Cyphers, Jim Glass
Automatic Speech Recognition and Understanding Workshop (ASRU), 2013
ASGARD: a Portable Architecture for Multilingual Dialogue Systems
Jingjing Liu, Panupong Pasupat, Scott Cyphers, Jim Glass
International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2013
paper
A Conversational Movie Search System Based on Conditional Random Fields
Jingjing Liu, Scott Cyphers, Panupong Pasupat, Ian Mcgraw, Jim Glass
Interspeech, 2012
paper

I usually take notes with a9, a Markdown editor I wrote before I learned about StackEdit. Here are some of my exported notes:

Sometimes I read stuff and then take notes. Sometimes I don't.

Resources for learning Chinese and Japanese:

A collection of math problems that have great significance to me:

My sister has been mailing me a birthday card every year, so I made a few for her:

Other fun stuff:

Double Crossword: a crossword that can be filled in either English or Thai! (Solutions)