Verification and implementation of language-based deception indicators in civil and criminal narratives

Joan Bachenko, Eileen Fitzpatrick, Michael Schonwetter

Research output: Chapter in Book/Report/Conference proceedingConference contributionResearchpeer-review

30 Citations (Scopus)

Abstract

Our goal is to use natural language processing to identify deceptive and non-deceptive passages in transcribed narratives. We begin by motivating an analysis of language-based deception that relies on specific linguistic indicators to discover deceptive statements. The indicator tags are assigned to a document using a mix of automated and manual methods. Once the tags are assigned, an interpreter automatically discriminates between deceptive and truthful statements based on tag densities. The texts used in our study come entirely from "real world" sources-criminal statements, police interrogations and legal testimony. The corpus was hand-tagged for the truth value of all propositions that could be externally verified as true or false. Classification and Regression Tree techniques suggest that the approach is feasible, with the model able to identify 74.9% of the T/F propositions correctly. Implementation of an automatic tagger with a large subset of tags performed well on test data, producing an average score of 68.6% recall and 85.3% precision when compared to the performance of human taggers on the same subset.

Original languageEnglish
Title of host publicationColing 2008 - 22nd International Conference on Computational Linguistics, Proceedings of the Conference
Pages41-48
Number of pages8
Volume1
StatePublished - 1 Dec 2008
Event22nd International Conference on Computational Linguistics, Coling 2008 - Manchester, United Kingdom
Duration: 18 Aug 200822 Aug 2008

Other

Other22nd International Conference on Computational Linguistics, Coling 2008
CountryUnited Kingdom
CityManchester
Period18/08/0822/08/08

Fingerprint

Law enforcement
Linguistics
narrative
Processing
language
interpreter
testimony
police
linguistics
regression
performance
Deception
Tag
Language
Values

Cite this

Bachenko, J., Fitzpatrick, E., & Schonwetter, M. (2008). Verification and implementation of language-based deception indicators in civil and criminal narratives. In Coling 2008 - 22nd International Conference on Computational Linguistics, Proceedings of the Conference (Vol. 1, pp. 41-48)
Bachenko, Joan ; Fitzpatrick, Eileen ; Schonwetter, Michael. / Verification and implementation of language-based deception indicators in civil and criminal narratives. Coling 2008 - 22nd International Conference on Computational Linguistics, Proceedings of the Conference. Vol. 1 2008. pp. 41-48
@inproceedings{77de8c19ddf34127978afdde040aefb8,
title = "Verification and implementation of language-based deception indicators in civil and criminal narratives",
abstract = "Our goal is to use natural language processing to identify deceptive and non-deceptive passages in transcribed narratives. We begin by motivating an analysis of language-based deception that relies on specific linguistic indicators to discover deceptive statements. The indicator tags are assigned to a document using a mix of automated and manual methods. Once the tags are assigned, an interpreter automatically discriminates between deceptive and truthful statements based on tag densities. The texts used in our study come entirely from {"}real world{"} sources-criminal statements, police interrogations and legal testimony. The corpus was hand-tagged for the truth value of all propositions that could be externally verified as true or false. Classification and Regression Tree techniques suggest that the approach is feasible, with the model able to identify 74.9{\%} of the T/F propositions correctly. Implementation of an automatic tagger with a large subset of tags performed well on test data, producing an average score of 68.6{\%} recall and 85.3{\%} precision when compared to the performance of human taggers on the same subset.",
author = "Joan Bachenko and Eileen Fitzpatrick and Michael Schonwetter",
year = "2008",
month = "12",
day = "1",
language = "English",
isbn = "9781905593446",
volume = "1",
pages = "41--48",
booktitle = "Coling 2008 - 22nd International Conference on Computational Linguistics, Proceedings of the Conference",

}

Bachenko, J, Fitzpatrick, E & Schonwetter, M 2008, Verification and implementation of language-based deception indicators in civil and criminal narratives. in Coling 2008 - 22nd International Conference on Computational Linguistics, Proceedings of the Conference. vol. 1, pp. 41-48, 22nd International Conference on Computational Linguistics, Coling 2008, Manchester, United Kingdom, 18/08/08.

Verification and implementation of language-based deception indicators in civil and criminal narratives. / Bachenko, Joan; Fitzpatrick, Eileen; Schonwetter, Michael.

Coling 2008 - 22nd International Conference on Computational Linguistics, Proceedings of the Conference. Vol. 1 2008. p. 41-48.

Research output: Chapter in Book/Report/Conference proceedingConference contributionResearchpeer-review

TY - GEN

T1 - Verification and implementation of language-based deception indicators in civil and criminal narratives

AU - Bachenko, Joan

AU - Fitzpatrick, Eileen

AU - Schonwetter, Michael

PY - 2008/12/1

Y1 - 2008/12/1

N2 - Our goal is to use natural language processing to identify deceptive and non-deceptive passages in transcribed narratives. We begin by motivating an analysis of language-based deception that relies on specific linguistic indicators to discover deceptive statements. The indicator tags are assigned to a document using a mix of automated and manual methods. Once the tags are assigned, an interpreter automatically discriminates between deceptive and truthful statements based on tag densities. The texts used in our study come entirely from "real world" sources-criminal statements, police interrogations and legal testimony. The corpus was hand-tagged for the truth value of all propositions that could be externally verified as true or false. Classification and Regression Tree techniques suggest that the approach is feasible, with the model able to identify 74.9% of the T/F propositions correctly. Implementation of an automatic tagger with a large subset of tags performed well on test data, producing an average score of 68.6% recall and 85.3% precision when compared to the performance of human taggers on the same subset.

AB - Our goal is to use natural language processing to identify deceptive and non-deceptive passages in transcribed narratives. We begin by motivating an analysis of language-based deception that relies on specific linguistic indicators to discover deceptive statements. The indicator tags are assigned to a document using a mix of automated and manual methods. Once the tags are assigned, an interpreter automatically discriminates between deceptive and truthful statements based on tag densities. The texts used in our study come entirely from "real world" sources-criminal statements, police interrogations and legal testimony. The corpus was hand-tagged for the truth value of all propositions that could be externally verified as true or false. Classification and Regression Tree techniques suggest that the approach is feasible, with the model able to identify 74.9% of the T/F propositions correctly. Implementation of an automatic tagger with a large subset of tags performed well on test data, producing an average score of 68.6% recall and 85.3% precision when compared to the performance of human taggers on the same subset.

UR - http://www.scopus.com/inward/record.url?scp=80053390928&partnerID=8YFLogxK

M3 - Conference contribution

SN - 9781905593446

VL - 1

SP - 41

EP - 48

BT - Coling 2008 - 22nd International Conference on Computational Linguistics, Proceedings of the Conference

ER -

Bachenko J, Fitzpatrick E, Schonwetter M. Verification and implementation of language-based deception indicators in civil and criminal narratives. In Coling 2008 - 22nd International Conference on Computational Linguistics, Proceedings of the Conference. Vol. 1. 2008. p. 41-48