“…The convergence criteria adopted for the EM algorithm is the termination tolerance on the log-likelihood function in (8), where the algorithm stops when the new guesses of parameters Θ * produce only minimal increments of the loglikelihood function given in (8), e.g. increments smaller than 10 −5 .…”
Section: Resultsmentioning
confidence: 99%
“…After the implementation of the BI System proposed in [7], with several improvements proposed in [8], [9] and [10], the current state of the art BI system for auditing SIAPE is based on an ontology indexation process via concept maps in order to detect irregularities on the payrolls, big data technologies such as Hadoop and Hbase for increasing the performance of the processing stage and a reimbursement tracking system for monitoring the payroll of federal employees who have to reimburse the Brazilian Treasury. Fig 1 shows the architecture of the current BI system.…”
Section: The Current Bi System Modelmentioning
confidence: 99%
“…The Human Resources Auditing Department (CGAUD) 1 , subordinated to the Brazilian Ministry of Planning, Budget and Management (MP) 2 , uses a specialized BI System to ascertain irregularities on the payrolls of the Brazilian federal government. The initial BI solution was presented in [7] and several improvements were proposed in [8], [9] and [10].…”
Section: Introductionmentioning
confidence: 99%
“…The current CGAUD BI system is entirely based on a knowledge-based approach for irregularity detection. The existing system uses audit trails built with ontological indexation via concept maps in order to detect inconsistencies [7,8,10]. The audit trails comprise a set of heuristics based on complex Brazilian federal legislation, which dictate the income of each public employee according to his career and his position in the public administration organization.…”
A Business Intelligence (BI) System employs tools from several areas of knowledge for the collection, integration and analysis of data to improve business decision making. The Brazilian Ministry of Planning, Budget and Management (MP) uses a BI System designed with the University of Brasília to ascertain irregularities on the payroll of the Brazilian federal government, performing audit trails on selected items and fields of the payroll database. This current auditing approach is entirely deterministic, since the audit trails look for previously known signatures of irregularities which are composed by means of an ontological method used to represent auditors concept maps. In this work, we propose to incorporate a statistical filter in this existing BI system in order to increase its performance in terms of processing speed and overall system responsiveness. The proposed statistical filter is based on a generative Gaussian Mixture Model (GMM) whose goal is to provide a complete stochastic model of the process, specially the latent probability density function of the generative mixture, and use that model to filter the most probable payrolls. Inserting this statistical filter as a pre-processing stage preceding the deterministic auditing showed to be effective in reducing the amount of data to be analyzed by the audit trails, despite the penalty fee intrinsically associated with stochastic models due to the false negative outcomes that are not further processed. In our approach, gains obtained with the proposed pre-processing stage overcome impacts from false negative outcomes.
“…The convergence criteria adopted for the EM algorithm is the termination tolerance on the log-likelihood function in (8), where the algorithm stops when the new guesses of parameters Θ * produce only minimal increments of the loglikelihood function given in (8), e.g. increments smaller than 10 −5 .…”
Section: Resultsmentioning
confidence: 99%
“…After the implementation of the BI System proposed in [7], with several improvements proposed in [8], [9] and [10], the current state of the art BI system for auditing SIAPE is based on an ontology indexation process via concept maps in order to detect irregularities on the payrolls, big data technologies such as Hadoop and Hbase for increasing the performance of the processing stage and a reimbursement tracking system for monitoring the payroll of federal employees who have to reimburse the Brazilian Treasury. Fig 1 shows the architecture of the current BI system.…”
Section: The Current Bi System Modelmentioning
confidence: 99%
“…The Human Resources Auditing Department (CGAUD) 1 , subordinated to the Brazilian Ministry of Planning, Budget and Management (MP) 2 , uses a specialized BI System to ascertain irregularities on the payrolls of the Brazilian federal government. The initial BI solution was presented in [7] and several improvements were proposed in [8], [9] and [10].…”
Section: Introductionmentioning
confidence: 99%
“…The current CGAUD BI system is entirely based on a knowledge-based approach for irregularity detection. The existing system uses audit trails built with ontological indexation via concept maps in order to detect inconsistencies [7,8,10]. The audit trails comprise a set of heuristics based on complex Brazilian federal legislation, which dictate the income of each public employee according to his career and his position in the public administration organization.…”
A Business Intelligence (BI) System employs tools from several areas of knowledge for the collection, integration and analysis of data to improve business decision making. The Brazilian Ministry of Planning, Budget and Management (MP) uses a BI System designed with the University of Brasília to ascertain irregularities on the payroll of the Brazilian federal government, performing audit trails on selected items and fields of the payroll database. This current auditing approach is entirely deterministic, since the audit trails look for previously known signatures of irregularities which are composed by means of an ontological method used to represent auditors concept maps. In this work, we propose to incorporate a statistical filter in this existing BI system in order to increase its performance in terms of processing speed and overall system responsiveness. The proposed statistical filter is based on a generative Gaussian Mixture Model (GMM) whose goal is to provide a complete stochastic model of the process, specially the latent probability density function of the generative mixture, and use that model to filter the most probable payrolls. Inserting this statistical filter as a pre-processing stage preceding the deterministic auditing showed to be effective in reducing the amount of data to be analyzed by the audit trails, despite the penalty fee intrinsically associated with stochastic models due to the false negative outcomes that are not further processed. In our approach, gains obtained with the proposed pre-processing stage overcome impacts from false negative outcomes.
“…The first step of the conception of a BI system is the identification and definition of the indicators required by the decision-making managers. In a previous work described in (Fernandes, Amaro, da Costa, Martins, Serrano and de Sousa Jr., 2012), it is shown how this process can be improved by the use of concept maps. This approach was applied to the concept of Audit Trails.…”
Business Intelligence (BI) systems are crucial for assisting the decision making processes of private and governmental institutions. The Human Resources Auditing Department (CGAUD) of the Brazilian Ministry of Planning, Budget and Management (MP) has been developing its own BI for auditing the payroll of all federal employees since 2010. Given that the monthly payroll is approximately 12.5 billion reais, the initial version of the proposed BI system in 2012 was able to audit approximately 1.5 billion reais. In this paper, we propose an improved BI system, which can deal with an increased volume of data, a greater amount of monitoring trails and a higher granularity of the final reports. As consequence, the total audit value has increased to approximately 5 billion reais. In addition, our new BI system has incorporated a Reimbursement Tracking System for monitoring the payroll of federal employees who have to reimburse the Brazilian government. Around 4.5 million reais are automatically monthly tracked by our new BI system. Our proposed BI system has been validated using the real environment of the MP and the results are compared to the previous BI system.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.