Modelo ProLab: Predictor, una propuesta para la identificación de patrones de fraude utilizando Inteligencia Artificial
Date
2024-07-08
Journal Title
Journal ISSN
Volume Title
Publisher
Pontificia Universidad Católica del Perú
Abstract
En el país, existe una gran cantidad de casos de fraude, esto también sucede en las empresas
del sector financiero. En la medida que se tiene un crecimiento económico en el país, la
mediana y pequeña empresa tiene la necesidad de invertir en sus negocios y para esto
requieren préstamos. En este ámbito, los bancos, cajas municipales, cajas rurales,
cooperativas, etc. desarrollan la intermediación financiera, que consta de trasladar los
recursos (dinero) del cliente ahorrista al cliente prestamista a una tasa mayor.
Los asesores de crédito obtienen préstamos para sus clientes, realizando diversas
evaluaciones y, según el nivel de riesgo, se realiza el préstamo. En varias empresas, esta
evaluación no sigue el orden correcto y para lograr las metas y recompensas los asesores no
siguen las evaluaciones correctas, lo que genera fraude a nivel crediticio. Los órganos de
control pertinentes no siempre se dan cuenta de esto. El ODS que se busca impactar es el
siguiente: ODS 16 “Paz, justicia e instituciones sólidas”; específicamente, con la meta 16.5
“Reducir considerablemente la corrupción y el soborno en todas sus formas”. Con este
proyecto se entregará un modelo de predicción de patrones de fraude, usando inteligencia
artificial, que evalúe de manera directa a los asesores de crédito, usando como origen y base
las centrales de riesgos externas (Infocorp, SBS, Sentinel, Sunarp, Reniec, antecedentes
policiales/penales/judiciales y migraciones) y fuentes de información internas (Siper-RRHH,
Core, SIG- Comisiones, Cosechas) y, en base al resultado de esta evaluación, dirigir las tareas
de auditoría acerca de la cartera de clientes de dichos colaboradores. En base a la realización
de la evaluación financiera del proyecto, se confirmó que es viable y rentable, con un VAN
de S/4’205,238.00 y un VANS de S/4’902,523.00, lo cual indicó que el proyecto recupera la
inversión y crea valor.
In the country, there are many cases of fraud, and this also happens in companies in the financial sector. As the country's economy grows, small and medium-sized enterprises need to invest in their businesses and therefore require loans. In this area, banks, municipal savings banks, rural savings banks, cooperatives, etc. carry out financial intermediation, which consists of transferring the resources (money) from the savings client to the lending client at a higher rate. Credit counsellors obtain loans for their clients by making various assessments and, depending on the level of risk, the loan is made. In several companies, this assessment does not follow the correct order and in order to achieve the goals and rewards the assessors do not follow the correct assessments, which leads to fraud at the credit level. This is not always noticed by the relevant control bodies. The SDG to be impacted is the following: SDG 16 "Peace, justice and strong institutions"; specifically, with target 16.5 "Significantly reduce corruption and bribery in all its forms". This project will provide a model for predicting fraud patterns, using artificial intelligence, which directly evaluates credit advisors, using external risk centres (Infocorp, SBS, Sentinel, Sunarp, Reniec, police/criminal/judicial records and migrations) and internal information sources (Siper-RR. HH, Core, SIG-commissions, harvests) as a source and basis, and, based on the result of this evaluation, direct the audit tasks on the client portfolio of these collaborators. Based on the financial evaluation of the project, it was confirmed that the project is viable and profitable, with an NPV of S/4'205,238.00 and an NPV of S/4'902,523.00, which indicated that the project recovers the investment and creates value.
In the country, there are many cases of fraud, and this also happens in companies in the financial sector. As the country's economy grows, small and medium-sized enterprises need to invest in their businesses and therefore require loans. In this area, banks, municipal savings banks, rural savings banks, cooperatives, etc. carry out financial intermediation, which consists of transferring the resources (money) from the savings client to the lending client at a higher rate. Credit counsellors obtain loans for their clients by making various assessments and, depending on the level of risk, the loan is made. In several companies, this assessment does not follow the correct order and in order to achieve the goals and rewards the assessors do not follow the correct assessments, which leads to fraud at the credit level. This is not always noticed by the relevant control bodies. The SDG to be impacted is the following: SDG 16 "Peace, justice and strong institutions"; specifically, with target 16.5 "Significantly reduce corruption and bribery in all its forms". This project will provide a model for predicting fraud patterns, using artificial intelligence, which directly evaluates credit advisors, using external risk centres (Infocorp, SBS, Sentinel, Sunarp, Reniec, police/criminal/judicial records and migrations) and internal information sources (Siper-RR. HH, Core, SIG-commissions, harvests) as a source and basis, and, based on the result of this evaluation, direct the audit tasks on the client portfolio of these collaborators. Based on the financial evaluation of the project, it was confirmed that the project is viable and profitable, with an NPV of S/4'205,238.00 and an NPV of S/4'902,523.00, which indicated that the project recovers the investment and creates value.
Description
Keywords
Fraudes financieros--Prevención--Perú, Inteligencia artificial--Riesgos de fraude
Citation
Endorsement
Review
Supplemented By
Referenced By
Creative Commons license
Except where otherwised noted, this item's license is described as info:eu-repo/semantics/openAccess