An Earth Mover’s Distance Based Graph Distance Metric For Financial Statements

Sander, Noels,; Vandermarliere, Benjamin; Bastiaensen, Ken; Bie, Tijl De

doi:10.1109/cifer52523.2022.9776204

Cited by 3 publications

(4 citation statements)

References 14 publications

(3 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Although the classification speed is significantly improved, classification accuracy is sacrificed. S. Noels et al [14] proposed a new graph distance metric based on the earth mover's distance to calculate the similarity between two enterprises, and the experimental results reached the expected results in many cases. However, the authors only focused on the balance sheet portion of the financial statements to construct the graph distance measurements, and many financial ratios were not utilized.…”

Section: Financial Fraud Detection Modelmentioning

confidence: 84%

Research on the Financial Data Fraud Detection of Chinese Listed Enterprises by Integrating Audit Opinions

2023

KSII TIIS

View full text Add to dashboard Cite

Financial fraud undermines the sustainable development of financial markets. Financial statements can be regarded as the key source of information to obtain the operating conditions of listed companies. Current research focuses more on mining financial digital data instead of looking into text data. However, text data can reveal emotional information, which is an important basis for detecting financial fraud. The audit opinion of the financial statement is especially the fair opinion of a certified public accountant on the quality of enterprise financial reports. Therefore, this research was carried out by using the data features of 4,153 listed companies' financial annual reports and audits of text opinions in the past six years, and the paper puts forward a financial fraud detection model integrating audit opinions. First, the financial data index database and audit opinion text database were built. Second, digitized audit opinions with deep learning Bert model was employed. Finally, both the extracted audit numerical characteristics and the financial numerical indicators were used as the training data of the LightGBM model. What is worth paying attention to is that the imbalanced distribution of sample labels is also one of the focuses of financial fraud research. To solve this problem, data enhancement and Focal Loss feature learning functions were used in data processing and model training respectively. The experimental results show that compared with the conventional financial fraud detection model, the performance of the proposed model is improved greatly, with Area Under the Curve (AUC) and Accuracy reaching 81.42% and 78.15%, respectively.

show abstract

Section: Financial Fraud Detection Modelmentioning

confidence: 84%

Research on the Financial Data Fraud Detection of Chinese Listed Enterprises by Integrating Audit Opinions

2023

KSII TIIS

View full text Add to dashboard Cite

show abstract

“…This research received funding from the Flemish Government, through Flanders Innovation & Entrepreneurship (VLAIO, project HBC.2020.2883) and from the Flemish Government under the “Onderzoeksprogramma Artificiële Intelligentie (AI) Vlaanderen” program. This is an extended and revised version of a preliminary conference proceeding that was presented at the 2022 IEEE Symposium on Computational Intelligence for Financial Engineering and Economics (CIFEr) (Noels et al, 2022). We would like to thank the Silverfin AI Team for their collaboration and support on this research project.…”

Section: Acknowledgmentsmentioning

confidence: 99%

An efficient graph‐based peer selection method for financial statements

Sander

Ridder²,

Viaene³

et al. 2023

Intell Sys Acc Fin Mgmt

Self Cite

View full text Add to dashboard Cite

SummaryComparing companies can be useful for various purposes. Despite the widespread use of industry classification systems as a peer selection standard, these have been criticized for various reasons. Financial statements, however, offer a promising alternative to such classification systems. They are standardized, widely available, and offer deep insights into the nature of the company. In this paper, we present a graph distance metric for financial statements using the earth mover's distance. When using the distance metric on real‐world tasks such as peer identification and industry classification, it shows promising results in terms of accuracy and computational efficiency.

show abstract

“…Reference [1] obtained the financial statement data from Silverfin, along with the financial ratios and industry activity codes. Reference [2] used the financial statement data of 1000 Belgian companies, with their disclosure published separately before 2019. Reference [5] used a dataset consisting of 71295 firm-years' financial statement data.…”

Section: Type I: Structured Data and Independentmentioning

confidence: 99%

“…Reference [1] presented a graph distance metric for financial statements using the earth mover's distance and tested the result with accuracy. Reference [2] changed the ledger accounts within a financial statement to a vertex-labelled tree, proposed a distance metric to measure the similarity between companies, and used a predictive method to verify the usefulness of the metric. Reference [13] used a deep learning method to encode the network graph of companies in a low-dimensional embedding space and validated the rationality of the method with accuracy-related indicators.…”

Section: Type Ii: Graph and Predictive Modelingmentioning

confidence: 99%

A Comparative Study on Measuring Similarity between Companies Based on Financial Statements

Zhang

2023

AEMR

View full text Add to dashboard Cite

It is important for financial researchers to measure similarity between companies. This passage classifies references in this area based on research objects and research methods. To be more specific, there are four research objects based on data structure and data connectivity. Research methods can also be classified into four types based on data form and modeling type. It can be concluded that there are academic blanks on models based on graph theory: no matter independent data or unstructured data should be used more wildly. Moreover, interconnected structured data should be used more often when vectors are introduced into predictive modeling.

show abstract

An Earth Mover’s Distance Based Graph Distance Metric For Financial Statements

Cited by 3 publications

References 14 publications

Research on the Financial Data Fraud Detection of Chinese Listed Enterprises by Integrating Audit Opinions

Research on the Financial Data Fraud Detection of Chinese Listed Enterprises by Integrating Audit Opinions

An efficient graph‐based peer selection method for financial statements

A Comparative Study on Measuring Similarity between Companies Based on Financial Statements

Contact Info

Product

Resources

About