A theory of linguistic individuality for authorship analysis / Andrea Nini.

Format:

Book

Author/Creator:

Nini, Andrea, author.

Series:

Cambridge elements, 2634-7334.

Cambridge elements. Elements in forensic linguistics, 2634-7334

Language:

English

Subjects (All):

Authorship--Data processing.

Authorship.

Natural language processing (Computer science)--Mathematical models.

Natural language processing (Computer science).

Physical Description:

1 online resource (96 pages) : illustrations (black and white), digital, PDF file(s).

Edition:

First edition.

Place of Publication:

Cambridge : Cambridge University Press, 2023.

Summary:

Authorship analysis is the process of determining who produced a questioned text by language analysis. Although there has been significant success in the performance of computational methods to solve this problem in recent years, these are often methods that are not amenable to interpretation. Authorship analysis is in all effects an area of computer science with very little linguistics or cognitive science. This Element introduces a Theory of Linguistic Individuality that, starting from basic notions of cognitive linguistics, establishes a formal framework for the mathematical modelling of language processing that is then applied to three computational experiments, including using the likelihood ratio framework. The results propose new avenues of research and a change of perspective in the way authorship analysis is currently carried out..

Contents:

Cover

Title page

Copyright page

A Theory of Linguistic Individuality for Authorship Analysis

Contents

Series Preface

1 Introduction

1.1 Authorship Analysis

1.2 Burrow's Delta

1.2.1 The Key Profiles Hypothesis

1.2.2 Explaining Authorship Analysis

1.3 Fundamentals of Cognitive Linguistics

1.3.1 Lexis and Grammar Are Inseparable

1.3.2 Language Is Processed in Chunks

1.3.3 Units Are Formed and Strengthened through Entrenchment and Conventionalisation

1.3.4 Grammar Is a Network

1.4 Idiolect and Linguistic Individuality

2 Formal Grounding for a Theory of Linguistic Individuality

2.1 Introduction

2.2 Basic Definitions

2.2.1 Linguistic Units

2.2.2 Grammar

2.2.3 Idiolect

2.2.4 Language and Variety

2.2.5 Text

2.3 The Principle of Linguistic Individuality

2.3.1 The Linguistic Fingerprint

2.4 The Statistical Approximation Hypothesis

3 Applying the Theory to Authorship Analysis

3.1 Grammar Induction

3.2 Measures of Similarity

3.3 A Reinterpretation of the N-gram Tracing Method

3.4 Experiment 1: The refcor Corpus

3.4.1 Individuality in n-grams

3.4.2 Maximising Known Sample Length

3.5 Experiment 2: The c50 Corpus

3.5.1 Previous Studies Using the c50 Corpus

3.5.2 Methodology

3.5.3 Results

3.6 Experiment 3: The Likelihood Ratio Framework

3.6.1 Calibration of Likelihood Ratios from Binary Coefficients

3.6.2 Experiment 3.1: c50 Background Data

3.6.3 Experiment 3.2: refcor Background Data

3.6.4 Conclusions

4 Discussion

4.1 The Cognitive Reality of n-grams and the Individuality Equilibrium

4.2 Evidence for the Statistical Approximation Hypothesis

4.3 Capturing an Individual Grammar in Full

4.4 The Evaluation of Similarity

4.5 An Individual's Grammar As a Behavioural Biometric

4.6 What About Style?.

4.7 A New Perspective on Authorship Analysis

5 Moving Forward

5.1 A Research Agenda for Future Work

5.2 Final Conclusions

5.3 Supplementary Material: Data and Code

References

Acknowledgments.

Notes:

Also issued in print: 2023.

Includes bibliographical references.

Description based on online resource; title from PDF title page (viewed on May 29, 2023).

ISBN:

9781108982177

1108982174

9781108985796

1108985793

9781108974851

1108974856

OCLC:

1379422508