DQI: Measuring Data Quality in NLP

by   Swaroop Mishra, et al.

Neural language models have achieved human level performance across several NLP datasets. However, recent studies have shown that these models are not truly learning the desired task; rather, their high performance is attributed to overfitting using spurious biases, which suggests that the capabilities of AI systems have been over-estimated. We introduce a generic formula for Data Quality Index (DQI) to help dataset creators create datasets free of such unwanted biases. We evaluate this formula using a recently proposed approach for adversarial filtering, AFLite. We propose a new data creation paradigm using DQI to create higher quality data. The data creation paradigm consists of several data visualizations to help data creators (i) understand the quality of data and (ii) visualize the impact of the created data instance on the overall quality. It also has a couple of automation methods to (i) assist data creators and (ii) make the model more robust to adversarial attacks. We use DQI along with these automation methods to renovate biased examples in SNLI. We show that models trained on the renovated SNLI dataset generalize better to out of distribution tasks. Renovation results in reduced model performance, exposing a large gap with respect to human performance. DQI systematically helps in creating harder benchmarks using active learning. Our work takes the process of dynamic dataset creation forward, wherein datasets evolve together with the evolving state of the art, therefore serving as a means of benchmarking the true progress of AI.


page 4

page 5

page 7

page 30

page 31

page 32

page 34

page 39


Adversarial Filters of Dataset Biases

Large neural models have demonstrated human-level performance on languag...

A Survey of Parameters Associated with the Quality of Benchmarks in NLP

Several benchmarks have been built with heavy investment in resources to...

Real-Time Visual Feedback to Guide Benchmark Creation: A Human-and-Metric-in-the-Loop Workflow

Recent research has shown that language models exploit `artifacts' in be...

Mapping global dynamics of benchmark creation and saturation in artificial intelligence

Benchmarks are crucial to measuring and steering progress in artificial ...

Improving Question Answering Model Robustness with Synthetic Adversarial Data Generation

Despite the availability of very large datasets and pretrained models, s...

Our Evaluation Metric Needs an Update to Encourage Generalization

Models that surpass human performance on several popular benchmarks disp...

Adversarially Constructed Evaluation Sets Are More Challenging, but May Not Be Fair

More capable language models increasingly saturate existing task benchma...

Please sign up or login with your details

Forgot password? Click here to reset