Fetching the paper…

Reading the bibliography…

2018

It was the training data pruning too!

Mudrakarta, Pramod Kaushik, Taly, Ankur, Sundararajan, Mukund et al.

Original
PDF
Post on X

Understand

We study the current best model (KDG) for question answering on tabular data evaluated over the WikiTableQuestions dataset.

Previous ablation studies performed against this model attributed the model's performance to certain aspects of its architecture.
In this paper, we find that the model's performance also crucially depends on a certain pruning of the data used to train the model.
Disabling the pruning step drops the accuracy of the model from 43.3% to 36.3%.

It was the training data pruning too! · Around

Built on

Machine learning as an experimental science
Dennis F. Kibler and Pat Langley. 1988 · 1988
Earlier work this paper cites.
Lambda dependency-based compositional semantics
Original
Percy Liang. 2013 · 2013
Earlier work this paper cites.
Compositional semantic parsing on semi-structured tables
Panupong Pasupat and Percy Liang. 2015 · 2015
Earlier work this paper cites.

Similar

Inferring logical forms from denotations
Original
Panupong Pasupat and Percy Liang. 2016 · 2016
Cited alongside, same era.
Neural multi-step reasoning for question answering on semi-structured tables
Original
Till Haug, Octavian-Eugen Ganea, and Paulina Grnarova. 2017 · 2017
Cited alongside, same era.

Then

Neural semantic parsing with type constraints for semi-structured tables
Jayant Krishnamurthy, Pradeep Dasigi, and Matt Gardner. 2017 · 2017
Later among the works it cites.
Learning a natural language interface with neural programmer
Arvind Neelakantan, Quoc V. Le, Martín Abadi, Andrew McCallum, and Dario Amodei. 2017 · 2017
Later among the works it cites.

Beyond the bibliography

alphaXiv searches the wider corpus for related work and actual follow-ups.

Open on alphaXiv

alphaXiv is searching for related work…