Named entity recognition in crime using machine learning approach

Hafedh Shabat, Nazlia Omar, Khmael Rahem

Research output: Contribution to journalArticle

9 Citations (Scopus)

Abstract

Most of the crimes committed today are reported on the Internet by news articles, blogs and social networking sites. With the increasing volume of crime information available on the web, a means to retrieve and exploit them and provide insight into the criminal behavior and networks must be determined to fight crime more efficiently and effectively. We believe that an electronic system must be designed for crime named entity recognition from the newspaper articles. Thus, this study designs and develops a crime named entity recognition based on machine learning approaches that extract nationalities, weapons, and crime locations in online crime documents. This study also collected a new corpus of crime and manually labeled them. A machine learning classification framework is proposed based on Naïve Bayes and SVM model in extracting nationalities, weapons, and crime location from online crime documents. To evaluate our model, a manually annotated data set was used, which was then validated by experiments. The results of the experiments showed that the developed techniques are promising.

Original languageEnglish
Pages (from-to)280-288
Number of pages9
JournalLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume8870
Publication statusPublished - 2014

Fingerprint

Named Entity Recognition
Crime
Learning systems
Machine Learning
Social Networking
Blogs
Bayes
Experiment
Electronics
Evaluate
Model
Social networking (online)
Experiments
Corpus
Framework
Internet

Keywords

  • Crime
  • Machine learning
  • Named entity recognition
  • Naïve bayes
  • Support vector machine

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

Cite this

@article{c0cb4b74c01d4ab2a1274c007af6c029,
title = "Named entity recognition in crime using machine learning approach",
abstract = "Most of the crimes committed today are reported on the Internet by news articles, blogs and social networking sites. With the increasing volume of crime information available on the web, a means to retrieve and exploit them and provide insight into the criminal behavior and networks must be determined to fight crime more efficiently and effectively. We believe that an electronic system must be designed for crime named entity recognition from the newspaper articles. Thus, this study designs and develops a crime named entity recognition based on machine learning approaches that extract nationalities, weapons, and crime locations in online crime documents. This study also collected a new corpus of crime and manually labeled them. A machine learning classification framework is proposed based on Na{\"i}ve Bayes and SVM model in extracting nationalities, weapons, and crime location from online crime documents. To evaluate our model, a manually annotated data set was used, which was then validated by experiments. The results of the experiments showed that the developed techniques are promising.",
keywords = "Crime, Machine learning, Named entity recognition, Na{\"i}ve bayes, Support vector machine",
author = "Hafedh Shabat and Nazlia Omar and Khmael Rahem",
year = "2014",
language = "English",
volume = "8870",
pages = "280--288",
journal = "Lecture Notes in Computer Science",
issn = "0302-9743",
publisher = "Springer Verlag",

}

TY - JOUR

T1 - Named entity recognition in crime using machine learning approach

AU - Shabat, Hafedh

AU - Omar, Nazlia

AU - Rahem, Khmael

PY - 2014

Y1 - 2014

N2 - Most of the crimes committed today are reported on the Internet by news articles, blogs and social networking sites. With the increasing volume of crime information available on the web, a means to retrieve and exploit them and provide insight into the criminal behavior and networks must be determined to fight crime more efficiently and effectively. We believe that an electronic system must be designed for crime named entity recognition from the newspaper articles. Thus, this study designs and develops a crime named entity recognition based on machine learning approaches that extract nationalities, weapons, and crime locations in online crime documents. This study also collected a new corpus of crime and manually labeled them. A machine learning classification framework is proposed based on Naïve Bayes and SVM model in extracting nationalities, weapons, and crime location from online crime documents. To evaluate our model, a manually annotated data set was used, which was then validated by experiments. The results of the experiments showed that the developed techniques are promising.

AB - Most of the crimes committed today are reported on the Internet by news articles, blogs and social networking sites. With the increasing volume of crime information available on the web, a means to retrieve and exploit them and provide insight into the criminal behavior and networks must be determined to fight crime more efficiently and effectively. We believe that an electronic system must be designed for crime named entity recognition from the newspaper articles. Thus, this study designs and develops a crime named entity recognition based on machine learning approaches that extract nationalities, weapons, and crime locations in online crime documents. This study also collected a new corpus of crime and manually labeled them. A machine learning classification framework is proposed based on Naïve Bayes and SVM model in extracting nationalities, weapons, and crime location from online crime documents. To evaluate our model, a manually annotated data set was used, which was then validated by experiments. The results of the experiments showed that the developed techniques are promising.

KW - Crime

KW - Machine learning

KW - Named entity recognition

KW - Naïve bayes

KW - Support vector machine

UR - http://www.scopus.com/inward/record.url?scp=84921325872&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84921325872&partnerID=8YFLogxK

M3 - Article

VL - 8870

SP - 280

EP - 288

JO - Lecture Notes in Computer Science

JF - Lecture Notes in Computer Science

SN - 0302-9743

ER -