Norm Conflict Identification Using a Convolutional Neural Network

João Paulo Aires; Felipe Meneguzzi

doi:10.1007/978-3-030-72376-7_1

Norm Conflict Identification Using a Convolutional Neural Network

João Paulo Aires, Felipe Meneguzzi

Pontifícia Universidade Católica do Rio Grande do Sul

Research output: Chapter in Book/Report/Conference proceeding › Chapter

Abstract

Contracts formally represent agreements between two or more parties as deontic statements or norms within their clauses. Norms may conflict between each other if not carefully designed, which may invalidate entire contracts. Human reviewers invest great effort to write conflict-free contracts that, for complex and long contracts, can be time-consuming and error-prone. In this work, we develop an approach to automate the identification of potential conflicts between norms in contracts. We build a two-phase approach that uses traditional machine learning together with a convolutional neural network to extract and compare norms in order to identify conflicts between them. Using a manually annotated and artificially generated set of conflicts as train and test set, our approach obtains 84% accuracy.

Original language	English
Title of host publication	Coordination, Organizations, Institutions, Norms, and Ethics for Governance of Multi-Agent Systems XIII
Subtitle of host publication	COIN COINE 2017 2020
Editors	A Aler Tubella, S Cranefield, C Frantz, F Meneguzzi, W Vasconcelos
Publisher	Springer
Volume	12298
ISBN (Print)	978-3-030-72375-0
DOIs	https://doi.org/10.1007/978-3-030-72376-7_1
Publication status	Published - 2 Apr 2021

Publication series

Name	Lecture Notes in Computer Science
Volume	12298

Bibliographical note

Acknowledgements

We gratefully thank Google Research Awards for Latin America for funding our project.

Keywords

Norms
Contracts
Deep learning
Natural language

Access to Document

10.1007/978-3-030-72376-7_1Licence: Unspecified

Cite this

Aires, J. P., & Meneguzzi, F. (2021). Norm Conflict Identification Using a Convolutional Neural Network. In A. Aler Tubella, S. Cranefield, C. Frantz, F. Meneguzzi, & W. Vasconcelos (Eds.), Coordination, Organizations, Institutions, Norms, and Ethics for Governance of Multi-Agent Systems XIII: COIN COINE 2017 2020 (Vol. 12298). (Lecture Notes in Computer Science; Vol. 12298). Springer . https://doi.org/10.1007/978-3-030-72376-7_1

Norm Conflict Identification Using a Convolutional Neural Network. / Aires, João Paulo; Meneguzzi, Felipe.
Coordination, Organizations, Institutions, Norms, and Ethics for Governance of Multi-Agent Systems XIII: COIN COINE 2017 2020. ed. / A Aler Tubella; S Cranefield; C Frantz; F Meneguzzi; W Vasconcelos. Vol. 12298 Springer , 2021. (Lecture Notes in Computer Science; Vol. 12298).

Research output: Chapter in Book/Report/Conference proceeding › Chapter

Aires, JP & Meneguzzi, F 2021, Norm Conflict Identification Using a Convolutional Neural Network. in A Aler Tubella, S Cranefield, C Frantz, F Meneguzzi & W Vasconcelos (eds), Coordination, Organizations, Institutions, Norms, and Ethics for Governance of Multi-Agent Systems XIII: COIN COINE 2017 2020. vol. 12298, Lecture Notes in Computer Science, vol. 12298, Springer . https://doi.org/10.1007/978-3-030-72376-7_1

Aires JP, Meneguzzi F. Norm Conflict Identification Using a Convolutional Neural Network. In Aler Tubella A, Cranefield S, Frantz C, Meneguzzi F, Vasconcelos W, editors, Coordination, Organizations, Institutions, Norms, and Ethics for Governance of Multi-Agent Systems XIII: COIN COINE 2017 2020. Vol. 12298. Springer . 2021. (Lecture Notes in Computer Science). doi: 10.1007/978-3-030-72376-7_1

Aires, João Paulo ; Meneguzzi, Felipe. / Norm Conflict Identification Using a Convolutional Neural Network. Coordination, Organizations, Institutions, Norms, and Ethics for Governance of Multi-Agent Systems XIII: COIN COINE 2017 2020. editor / A Aler Tubella ; S Cranefield ; C Frantz ; F Meneguzzi ; W Vasconcelos. Vol. 12298 Springer , 2021. (Lecture Notes in Computer Science).

@inbook{63ac103ff689496c831193fbc2b67ce1,

title = "Norm Conflict Identification Using a Convolutional Neural Network",

abstract = "Contracts formally represent agreements between two or more parties as deontic statements or norms within their clauses. Norms may conflict between each other if not carefully designed, which may invalidate entire contracts. Human reviewers invest great effort to write conflict-free contracts that, for complex and long contracts, can be time-consuming and error-prone. In this work, we develop an approach to automate the identification of potential conflicts between norms in contracts. We build a two-phase approach that uses traditional machine learning together with a convolutional neural network to extract and compare norms in order to identify conflicts between them. Using a manually annotated and artificially generated set of conflicts as train and test set, our approach obtains 84% accuracy.",

keywords = "Norms, Contracts, Deep learning, Natural language",

author = "Aires, {Jo{\~a}o Paulo} and Felipe Meneguzzi",

note = "Acknowledgements We gratefully thank Google Research Awards for Latin America for funding our project. ",

year = "2021",

month = apr,

day = "2",

doi = "10.1007/978-3-030-72376-7_1",

language = "English",

isbn = "978-3-030-72375-0",

volume = "12298",

series = "Lecture Notes in Computer Science",

publisher = "Springer ",

editor = "{Aler Tubella}, A and S Cranefield and C Frantz and F Meneguzzi and W Vasconcelos",

booktitle = "Coordination, Organizations, Institutions, Norms, and Ethics for Governance of Multi-Agent Systems XIII",

}

TY - CHAP

T1 - Norm Conflict Identification Using a Convolutional Neural Network

AU - Aires, João Paulo

AU - Meneguzzi, Felipe

N1 - Acknowledgements We gratefully thank Google Research Awards for Latin America for funding our project.

PY - 2021/4/2

Y1 - 2021/4/2

N2 - Contracts formally represent agreements between two or more parties as deontic statements or norms within their clauses. Norms may conflict between each other if not carefully designed, which may invalidate entire contracts. Human reviewers invest great effort to write conflict-free contracts that, for complex and long contracts, can be time-consuming and error-prone. In this work, we develop an approach to automate the identification of potential conflicts between norms in contracts. We build a two-phase approach that uses traditional machine learning together with a convolutional neural network to extract and compare norms in order to identify conflicts between them. Using a manually annotated and artificially generated set of conflicts as train and test set, our approach obtains 84% accuracy.

AB - Contracts formally represent agreements between two or more parties as deontic statements or norms within their clauses. Norms may conflict between each other if not carefully designed, which may invalidate entire contracts. Human reviewers invest great effort to write conflict-free contracts that, for complex and long contracts, can be time-consuming and error-prone. In this work, we develop an approach to automate the identification of potential conflicts between norms in contracts. We build a two-phase approach that uses traditional machine learning together with a convolutional neural network to extract and compare norms in order to identify conflicts between them. Using a manually annotated and artificially generated set of conflicts as train and test set, our approach obtains 84% accuracy.

KW - Norms

KW - Contracts

KW - Deep learning

KW - Natural language

UR - https://doi.org/10.1007/978-3-030-72376-7_1

U2 - 10.1007/978-3-030-72376-7_1

DO - 10.1007/978-3-030-72376-7_1

M3 - Chapter

SN - 978-3-030-72375-0

VL - 12298

T3 - Lecture Notes in Computer Science

BT - Coordination, Organizations, Institutions, Norms, and Ethics for Governance of Multi-Agent Systems XIII

A2 - Aler Tubella, A

A2 - Cranefield, S

A2 - Frantz, C

A2 - Meneguzzi, F

A2 - Vasconcelos, W

PB - Springer

ER -

Norm Conflict Identification Using a Convolutional Neural Network

Abstract

Publication series

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this