Knowledge Base Ontology Building for Fraud Detection Using Topic Modeling

Girija Attigeri, M. M. Manohara Pai, Radhika M. Pai, Rahul Kulkarni

Research output: Contribution to journalConference article

Abstract

Moving towards the digitization and cashless economy tests the existing IT infrastructure for security and fraud controls substantially. Transition from traditional to cashless economy requires to banks to have more secure system to fight fraud. To understand and transform the needs for more secure banking system it is necessary to understand the domain of fraud and create knowledge base for fraud. It helps bridge the gap between business level and IT levels of banking. So that anti-fraud regulations could be automatically imbibed in the system. Hence the paper focuses on analyzing existing fraud case documentations and understand the significant terms involved in the fraud. For this TF-IDF weighting, topic modeling with LDA is used for identifying the group of words (topic) representing particular type of fraud. Using these knowledge base ontology is extracted which can be used for building fraud detection system. Experiment is performed on extracted fraud documents and ontology is built using the latent topics identified.

Original languageEnglish
Pages (from-to)369-376
Number of pages8
JournalProcedia Computer Science
Volume135
DOIs
Publication statusPublished - 01-01-2018
Event3rd International Conference on Computer Science and Computational Intelligence, ICCSCI 2018 - Tangerang, Indonesia
Duration: 07-09-201808-09-2018

Fingerprint

Ontology
Analog to digital conversion
Industry
Experiments

All Science Journal Classification (ASJC) codes

  • Computer Science(all)

Cite this

@article{61dd29e60a054d6c97cabb149521ddff,
title = "Knowledge Base Ontology Building for Fraud Detection Using Topic Modeling",
abstract = "Moving towards the digitization and cashless economy tests the existing IT infrastructure for security and fraud controls substantially. Transition from traditional to cashless economy requires to banks to have more secure system to fight fraud. To understand and transform the needs for more secure banking system it is necessary to understand the domain of fraud and create knowledge base for fraud. It helps bridge the gap between business level and IT levels of banking. So that anti-fraud regulations could be automatically imbibed in the system. Hence the paper focuses on analyzing existing fraud case documentations and understand the significant terms involved in the fraud. For this TF-IDF weighting, topic modeling with LDA is used for identifying the group of words (topic) representing particular type of fraud. Using these knowledge base ontology is extracted which can be used for building fraud detection system. Experiment is performed on extracted fraud documents and ontology is built using the latent topics identified.",
author = "Girija Attigeri and {Manohara Pai}, {M. M.} and Pai, {Radhika M.} and Rahul Kulkarni",
year = "2018",
month = "1",
day = "1",
doi = "10.1016/j.procs.2018.08.186",
language = "English",
volume = "135",
pages = "369--376",
journal = "Procedia Computer Science",
issn = "1877-0509",
publisher = "Elsevier BV",

}

Knowledge Base Ontology Building for Fraud Detection Using Topic Modeling. / Attigeri, Girija; Manohara Pai, M. M.; Pai, Radhika M.; Kulkarni, Rahul.

In: Procedia Computer Science, Vol. 135, 01.01.2018, p. 369-376.

Research output: Contribution to journalConference article

TY - JOUR

T1 - Knowledge Base Ontology Building for Fraud Detection Using Topic Modeling

AU - Attigeri, Girija

AU - Manohara Pai, M. M.

AU - Pai, Radhika M.

AU - Kulkarni, Rahul

PY - 2018/1/1

Y1 - 2018/1/1

N2 - Moving towards the digitization and cashless economy tests the existing IT infrastructure for security and fraud controls substantially. Transition from traditional to cashless economy requires to banks to have more secure system to fight fraud. To understand and transform the needs for more secure banking system it is necessary to understand the domain of fraud and create knowledge base for fraud. It helps bridge the gap between business level and IT levels of banking. So that anti-fraud regulations could be automatically imbibed in the system. Hence the paper focuses on analyzing existing fraud case documentations and understand the significant terms involved in the fraud. For this TF-IDF weighting, topic modeling with LDA is used for identifying the group of words (topic) representing particular type of fraud. Using these knowledge base ontology is extracted which can be used for building fraud detection system. Experiment is performed on extracted fraud documents and ontology is built using the latent topics identified.

AB - Moving towards the digitization and cashless economy tests the existing IT infrastructure for security and fraud controls substantially. Transition from traditional to cashless economy requires to banks to have more secure system to fight fraud. To understand and transform the needs for more secure banking system it is necessary to understand the domain of fraud and create knowledge base for fraud. It helps bridge the gap between business level and IT levels of banking. So that anti-fraud regulations could be automatically imbibed in the system. Hence the paper focuses on analyzing existing fraud case documentations and understand the significant terms involved in the fraud. For this TF-IDF weighting, topic modeling with LDA is used for identifying the group of words (topic) representing particular type of fraud. Using these knowledge base ontology is extracted which can be used for building fraud detection system. Experiment is performed on extracted fraud documents and ontology is built using the latent topics identified.

UR - http://www.scopus.com/inward/record.url?scp=85053105297&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85053105297&partnerID=8YFLogxK

U2 - 10.1016/j.procs.2018.08.186

DO - 10.1016/j.procs.2018.08.186

M3 - Conference article

VL - 135

SP - 369

EP - 376

JO - Procedia Computer Science

JF - Procedia Computer Science

SN - 1877-0509

ER -