Leveraging Financial Social Media Data for Corporate Fraud Detection

Wei Dong, Shaoyi Liao, Zhongju Zhang

Research output: Contribution to journalArticle

6 Citations (Scopus)

Abstract

Corporate fraud can lead to significant financial losses and cause immeasurable damage to investor confidence and the overall economy. Detection of such frauds is a time-consuming and challenging task. Traditionally, researchers have been relying on financial data and/or textual content from financial statements to detect corporate fraud. Guided by systemic functional linguistics (SFL) theory, we propose an analytic framework that taps into unstructured data from financial social media platforms to assess the risk of corporate fraud. We assemble a unique data set including 64 fraudulent firms and a matched sample of 64 nonfraudulent firms, as well as the social media data prior to the firm’s alleged fraud violation in Accounting and Auditing Enforcement Releases (AAERs). Our framework automatically extracts signals such as sentiment features, emotion features, topic features, lexical features, and social network features, which are then fed into machine learning classifiers for fraud detection. We evaluate and compare the performance of our algorithm against baseline approaches using only financial ratios and language-based features respectively. We further validate the robustness of our algorithm by detecting leaked information and rumors, testing the algorithm on a new data set, and conducting an applicability check. Our results demonstrate the value of financial social media data and serve as a proof of concept of using such data to complement traditional fraud detection methods.

Original languageEnglish (US)
Pages (from-to)461-487
Number of pages27
JournalJournal of Management Information Systems
Volume35
Issue number2
DOIs
StatePublished - Apr 3 2018

Fingerprint

Linguistics
Learning systems
Classifiers
Testing
Fraud
Social media
Fraud detection
Confidence
Violations
Social networks
Emotion
Damage
Financial ratios
Language
Robustness
Rumor
Financial statements
Machine learning
Financial data
Classifier

ASJC Scopus subject areas

  • Management Information Systems
  • Computer Science Applications
  • Management Science and Operations Research
  • Information Systems and Management

Cite this

Leveraging Financial Social Media Data for Corporate Fraud Detection. / Dong, Wei; Liao, Shaoyi; Zhang, Zhongju.

In: Journal of Management Information Systems, Vol. 35, No. 2, 03.04.2018, p. 461-487.

Research output: Contribution to journalArticle

@article{6a080dd37f3a4c0182fbcd2f41d2d65d,
title = "Leveraging Financial Social Media Data for Corporate Fraud Detection",
abstract = "Corporate fraud can lead to significant financial losses and cause immeasurable damage to investor confidence and the overall economy. Detection of such frauds is a time-consuming and challenging task. Traditionally, researchers have been relying on financial data and/or textual content from financial statements to detect corporate fraud. Guided by systemic functional linguistics (SFL) theory, we propose an analytic framework that taps into unstructured data from financial social media platforms to assess the risk of corporate fraud. We assemble a unique data set including 64 fraudulent firms and a matched sample of 64 nonfraudulent firms, as well as the social media data prior to the firm’s alleged fraud violation in Accounting and Auditing Enforcement Releases (AAERs). Our framework automatically extracts signals such as sentiment features, emotion features, topic features, lexical features, and social network features, which are then fed into machine learning classifiers for fraud detection. We evaluate and compare the performance of our algorithm against baseline approaches using only financial ratios and language-based features respectively. We further validate the robustness of our algorithm by detecting leaked information and rumors, testing the algorithm on a new data set, and conducting an applicability check. Our results demonstrate the value of financial social media data and serve as a proof of concept of using such data to complement traditional fraud detection methods.",
author = "Wei Dong and Shaoyi Liao and Zhongju Zhang",
year = "2018",
month = "4",
day = "3",
doi = "10.1080/07421222.2018.1451954",
language = "English (US)",
volume = "35",
pages = "461--487",
journal = "Journal of Management Information Systems",
issn = "0742-1222",
publisher = "M.E. Sharpe Inc.",
number = "2",

}

TY - JOUR

T1 - Leveraging Financial Social Media Data for Corporate Fraud Detection

AU - Dong, Wei

AU - Liao, Shaoyi

AU - Zhang, Zhongju

PY - 2018/4/3

Y1 - 2018/4/3

N2 - Corporate fraud can lead to significant financial losses and cause immeasurable damage to investor confidence and the overall economy. Detection of such frauds is a time-consuming and challenging task. Traditionally, researchers have been relying on financial data and/or textual content from financial statements to detect corporate fraud. Guided by systemic functional linguistics (SFL) theory, we propose an analytic framework that taps into unstructured data from financial social media platforms to assess the risk of corporate fraud. We assemble a unique data set including 64 fraudulent firms and a matched sample of 64 nonfraudulent firms, as well as the social media data prior to the firm’s alleged fraud violation in Accounting and Auditing Enforcement Releases (AAERs). Our framework automatically extracts signals such as sentiment features, emotion features, topic features, lexical features, and social network features, which are then fed into machine learning classifiers for fraud detection. We evaluate and compare the performance of our algorithm against baseline approaches using only financial ratios and language-based features respectively. We further validate the robustness of our algorithm by detecting leaked information and rumors, testing the algorithm on a new data set, and conducting an applicability check. Our results demonstrate the value of financial social media data and serve as a proof of concept of using such data to complement traditional fraud detection methods.

AB - Corporate fraud can lead to significant financial losses and cause immeasurable damage to investor confidence and the overall economy. Detection of such frauds is a time-consuming and challenging task. Traditionally, researchers have been relying on financial data and/or textual content from financial statements to detect corporate fraud. Guided by systemic functional linguistics (SFL) theory, we propose an analytic framework that taps into unstructured data from financial social media platforms to assess the risk of corporate fraud. We assemble a unique data set including 64 fraudulent firms and a matched sample of 64 nonfraudulent firms, as well as the social media data prior to the firm’s alleged fraud violation in Accounting and Auditing Enforcement Releases (AAERs). Our framework automatically extracts signals such as sentiment features, emotion features, topic features, lexical features, and social network features, which are then fed into machine learning classifiers for fraud detection. We evaluate and compare the performance of our algorithm against baseline approaches using only financial ratios and language-based features respectively. We further validate the robustness of our algorithm by detecting leaked information and rumors, testing the algorithm on a new data set, and conducting an applicability check. Our results demonstrate the value of financial social media data and serve as a proof of concept of using such data to complement traditional fraud detection methods.

UR - http://www.scopus.com/inward/record.url?scp=85047254879&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85047254879&partnerID=8YFLogxK

U2 - 10.1080/07421222.2018.1451954

DO - 10.1080/07421222.2018.1451954

M3 - Article

VL - 35

SP - 461

EP - 487

JO - Journal of Management Information Systems

JF - Journal of Management Information Systems

SN - 0742-1222

IS - 2

ER -