Martha Palmer, Daniel Gildea, and Nianwen Xue
2010
Spoken Dialogue Systems
Kristiina Jokinen and Michael McTear
2009
Introduction to Chinese Natural Language Processing
Kam-Fai Wong, Wenjie Li, Ruifeng Xu, and Zheng-sheng Zhang
2009
Introduction to Linguistic Annotation and Text Analytics
Graham Wilcock
2009
Dependency Parsing
Sandra Kübler, Ryan McDonald, and Joakim Nivre
2009
Statistical Language Models for Information Retrieval
ChengXiang Zhai
2008
Copyright © 2020 by Morgan & Claypool
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means—electronic, mechanical, photocopy, recording, or any other except for brief quotations in printed reviews, without the prior permission of the publisher.
Statistical Significance Testing for Natural Language Processing
Rotem Dror, Lotem Peled-Cohen, Segev Shlomov, and Roi Reichart
www.morganclaypool.com
ISBN: 9781681737959 paperback
ISBN: 9781681737966 ebook
ISBN: 9781681738307 epub
ISBN: 9781681737973 hardcover
DOI 10.2200/S00994ED1V01Y202002HLT045
A Publication in the Morgan & Claypool Publishers series
SYNTHESIS LECTURES ON HUMAN LANGUAGE TECHNOLOGIES
Lecture #45
Series Editor: Graeme Hirst, University of Toronto
Series ISSN
Print 1947-4040 Electronic 1947-4059
Statistical Significance Testing for Natural Language Processing
Rotem Dror, Lotem Peled-Cohen, Segev Shlomov, and Roi Reichart
Technion – Israel Institute of Technology
SYNTHESIS LECTURES ON HUMAN LANGUAGE TECHNOLOGIES #45
ABSTRACT
Data-driven experimental analysis has become the main evaluation tool of Natural Language Processing (NLP) algorithms. In fact, in the last decade, it has become rare to see an NLP paper, particularly one that proposes a new algorithm, that does not include extensive experimental analysis, and the number of involved tasks, datasets, domains, and languages is constantly growing. This emphasis on empirical results highlights the role of statistical significance testing in NLP research: If we, as a community, rely on empirical evaluation to validate our hypotheses and reveal the correct language processing mechanisms, we better be sure that our results are not coincidental.
The goal of this book is to discuss the main aspects of statistical significance testing in NLP. Our guiding assumption throughout the book is that the basic question NLP researchers and engineers deal with is whether or not one algorithm can be considered better than another one. This question drives the field forward as it allows the constant progress of developing better technology for language processing challenges. In practice, researchers and engineers would like to draw the right conclusion from a limited set of experiments, and this conclusion should hold for other experiments with datasets they do not have at their disposal or that they cannot perform due to limited time and resources. The book hence discusses the opportunities and challenges in using statistical significance testing in NLP, from the point of view of experimental comparison between two algorithms. We cover topics such as choosing an appropriate significance test for the major NLP tasks, dealing with the unique aspects of significance testing for non-convex deep neural networks, accounting for a large number of comparisons between two NLP algorithms in a statistically valid manner (multiple hypothesis testing), and, finally, the unique challenges yielded by the nature of the data and practices of the field.
KEYWORDS
Natural Language Processing, statistics, statistical significance, hypothesis testing, algorithm comparison, deep neural network models, replicability analysis
Contents
2 Statistical Hypothesis Testing
2.2 P-Value in the World of NLP
3 Statistical Significance Tests
4 Statistical Significance in NLP
4.1 NLP Tasks and Evaluation Measures
4.2 Decision Tree for Significance Test Selection
4.3 Matching Between Evaluation Measures and Statistical Significance Tests
4.4 Significance with Large Test Samples
5.1 Performance Variance in Deep Neural Network Models
5.2 A Deep Neural Network Comparison Framework
5.3 Existing Methods for Deep Neural Network Comparison
5.4 Almost Stochastic Dominance