A hybrid deep learning system of CNN and LRCN to detect cyberbullying from SNS comments

Seok Jun Bu, Sung-Bae Cho

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

The cyberbullying is becoming a significant social issue, in proportion to the proliferation of Social Network Service (SNS). The cyberbullying commentaries can be categorized into syntactic and semantic subsets. In this paper, we propose an ensemble method of the two deep learning models: One is character-level CNN which captures low-level syntactic information from the sequence of characters and is robust to noise using the transfer learning. The other is word-level LRCN which captures high-level semantic information from the sequence of words, complementing the CNN model. Empirical results show that the performance of the ensemble method is significantly enhanced, outperforming the state-of-the-art methods for detecting cyberbullying comment. The model is analyzed by t-SNE algorithm to investigate the mutually cooperative relations between syntactic and semantic models.

LanguageEnglish
Title of host publicationHybrid Artificial Intelligent Systems - 13th International Conference, HAIS 2018, Proceedings
EditorsAlvaro Herrero, Hector Quintian, Jose Antonio Saez, Emilio Corchado, Francisco Javier de Cos Juez, Jose Ramon Villar, Enrique A. de la Cal
PublisherSpringer Verlag
Pages561-572
Number of pages12
ISBN (Print)9783319926384
DOIs
Publication statusPublished - 2018 Jan 1
Event13th International Conference on Hybrid Artificial Intelligent Systems, HAIS 2018 - Oviedo, Spain
Duration: 2018 Jun 202018 Jun 22

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume10870 LNAI
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other13th International Conference on Hybrid Artificial Intelligent Systems, HAIS 2018
CountrySpain
CityOviedo
Period18/6/2018/6/22

Fingerprint

Hybrid Learning
Learning Systems
Hybrid Systems
Social Networks
Learning systems
Syntactics
Ensemble Methods
Semantics
Transfer Learning
Proliferation
Model
Proportion
Subset
Deep learning
Syntax
Character

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Bu, S. J., & Cho, S-B. (2018). A hybrid deep learning system of CNN and LRCN to detect cyberbullying from SNS comments. In A. Herrero, H. Quintian, J. Antonio Saez, E. Corchado, F. J. de Cos Juez, J. R. Villar, & E. A. de la Cal (Eds.), Hybrid Artificial Intelligent Systems - 13th International Conference, HAIS 2018, Proceedings (pp. 561-572). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10870 LNAI). Springer Verlag. https://doi.org/10.1007/978-3-319-92639-1_47
Bu, Seok Jun ; Cho, Sung-Bae. / A hybrid deep learning system of CNN and LRCN to detect cyberbullying from SNS comments. Hybrid Artificial Intelligent Systems - 13th International Conference, HAIS 2018, Proceedings. editor / Alvaro Herrero ; Hector Quintian ; Jose Antonio Saez ; Emilio Corchado ; Francisco Javier de Cos Juez ; Jose Ramon Villar ; Enrique A. de la Cal. Springer Verlag, 2018. pp. 561-572 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{ab4bfec1a8f245ff9575b8e6c66443b3,
title = "A hybrid deep learning system of CNN and LRCN to detect cyberbullying from SNS comments",
abstract = "The cyberbullying is becoming a significant social issue, in proportion to the proliferation of Social Network Service (SNS). The cyberbullying commentaries can be categorized into syntactic and semantic subsets. In this paper, we propose an ensemble method of the two deep learning models: One is character-level CNN which captures low-level syntactic information from the sequence of characters and is robust to noise using the transfer learning. The other is word-level LRCN which captures high-level semantic information from the sequence of words, complementing the CNN model. Empirical results show that the performance of the ensemble method is significantly enhanced, outperforming the state-of-the-art methods for detecting cyberbullying comment. The model is analyzed by t-SNE algorithm to investigate the mutually cooperative relations between syntactic and semantic models.",
author = "Bu, {Seok Jun} and Sung-Bae Cho",
year = "2018",
month = "1",
day = "1",
doi = "10.1007/978-3-319-92639-1_47",
language = "English",
isbn = "9783319926384",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "561--572",
editor = "Alvaro Herrero and Hector Quintian and {Antonio Saez}, Jose and Emilio Corchado and {de Cos Juez}, {Francisco Javier} and Villar, {Jose Ramon} and {de la Cal}, {Enrique A.}",
booktitle = "Hybrid Artificial Intelligent Systems - 13th International Conference, HAIS 2018, Proceedings",
address = "Germany",

}

Bu, SJ & Cho, S-B 2018, A hybrid deep learning system of CNN and LRCN to detect cyberbullying from SNS comments. in A Herrero, H Quintian, J Antonio Saez, E Corchado, FJ de Cos Juez, JR Villar & EA de la Cal (eds), Hybrid Artificial Intelligent Systems - 13th International Conference, HAIS 2018, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10870 LNAI, Springer Verlag, pp. 561-572, 13th International Conference on Hybrid Artificial Intelligent Systems, HAIS 2018, Oviedo, Spain, 18/6/20. https://doi.org/10.1007/978-3-319-92639-1_47

A hybrid deep learning system of CNN and LRCN to detect cyberbullying from SNS comments. / Bu, Seok Jun; Cho, Sung-Bae.

Hybrid Artificial Intelligent Systems - 13th International Conference, HAIS 2018, Proceedings. ed. / Alvaro Herrero; Hector Quintian; Jose Antonio Saez; Emilio Corchado; Francisco Javier de Cos Juez; Jose Ramon Villar; Enrique A. de la Cal. Springer Verlag, 2018. p. 561-572 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10870 LNAI).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - A hybrid deep learning system of CNN and LRCN to detect cyberbullying from SNS comments

AU - Bu, Seok Jun

AU - Cho, Sung-Bae

PY - 2018/1/1

Y1 - 2018/1/1

N2 - The cyberbullying is becoming a significant social issue, in proportion to the proliferation of Social Network Service (SNS). The cyberbullying commentaries can be categorized into syntactic and semantic subsets. In this paper, we propose an ensemble method of the two deep learning models: One is character-level CNN which captures low-level syntactic information from the sequence of characters and is robust to noise using the transfer learning. The other is word-level LRCN which captures high-level semantic information from the sequence of words, complementing the CNN model. Empirical results show that the performance of the ensemble method is significantly enhanced, outperforming the state-of-the-art methods for detecting cyberbullying comment. The model is analyzed by t-SNE algorithm to investigate the mutually cooperative relations between syntactic and semantic models.

AB - The cyberbullying is becoming a significant social issue, in proportion to the proliferation of Social Network Service (SNS). The cyberbullying commentaries can be categorized into syntactic and semantic subsets. In this paper, we propose an ensemble method of the two deep learning models: One is character-level CNN which captures low-level syntactic information from the sequence of characters and is robust to noise using the transfer learning. The other is word-level LRCN which captures high-level semantic information from the sequence of words, complementing the CNN model. Empirical results show that the performance of the ensemble method is significantly enhanced, outperforming the state-of-the-art methods for detecting cyberbullying comment. The model is analyzed by t-SNE algorithm to investigate the mutually cooperative relations between syntactic and semantic models.

UR - http://www.scopus.com/inward/record.url?scp=85048857617&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85048857617&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-92639-1_47

DO - 10.1007/978-3-319-92639-1_47

M3 - Conference contribution

SN - 9783319926384

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 561

EP - 572

BT - Hybrid Artificial Intelligent Systems - 13th International Conference, HAIS 2018, Proceedings

A2 - Herrero, Alvaro

A2 - Quintian, Hector

A2 - Antonio Saez, Jose

A2 - Corchado, Emilio

A2 - de Cos Juez, Francisco Javier

A2 - Villar, Jose Ramon

A2 - de la Cal, Enrique A.

PB - Springer Verlag

ER -

Bu SJ, Cho S-B. A hybrid deep learning system of CNN and LRCN to detect cyberbullying from SNS comments. In Herrero A, Quintian H, Antonio Saez J, Corchado E, de Cos Juez FJ, Villar JR, de la Cal EA, editors, Hybrid Artificial Intelligent Systems - 13th International Conference, HAIS 2018, Proceedings. Springer Verlag. 2018. p. 561-572. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-319-92639-1_47