Authors
Advisor(s)
Abstract(s)
Neste projeto era pretendida análise e investigação de dados organizacionais com vista a identificar áreas de melhoria e a corrigir. Assim, este foi desenvolvido com recurso a uma análise exploratória extensa e com base em indicadores de desempenho. Foram desenvolvidos, também, dashboards com base nas análises efetuadas e conforme as necessidades do utilizador, tendo em conta aspetos como a complexidade visual e a pertinência dos dados apresentados. Por outro lado, foi estudada a implementação de um algoritmo de previsão de vendas, analisando métodos como o método de Holt-Winters Aditivo, bem como o método SARIMA, dada a avaliação da estacionariedade da série temporal considerada e a presença de sazonalidade nos dados. Por fim, tendo isto foi possível identificar algumas falhas relativas à inserção dos dados considerados, que é efetuada no ERP organizacional. Por outro lado, verificou-se, também, que existe alguma falta de planeamento de produção, provocando sobrecarga em determinados centros de carga. Estes resultados são importantes por um lado porque o conhecimento que se pode adquirir através da análise de dados está inteiramente ligado à qualidade dos mesmos, pelo que deve ser promovida uma cultura organizacional que valorize este tipo de atividades de controlo do desempenho da organização, já que é delas que este depende. Por outro lado, a identificação da sobrecarga de centros de trabalho permite alertar os responsáveis pelo planeamento da produção sobre esta falha, uma vez que condiciona tanto a qualidade e prazos de entrega dos artigos ou serviços, bem como a motivação dos colaboradores.
In this project, the aim was to analyze and investigate organizational data in order to identify areas for improvement and adjustments. Therefore, extensive exploratory analysis was conducted, based on performance indicators. Additionally, dashboards were developed to meet user needs, taking into account aspects such as visual complexity and the relevance of the presented data. Furthermore, the implementation of a sales forecasting algorithm was studied, considering methods such as the Additive Holt-Winters method and the SARIMA method. This analysis was performed considering the stationarity evaluation of the time series data and the presence of seasonality in the data. Finally, based on these findings, some flaws were identified regarding the data input process in the organizational ERP. Moreover, it was observed that there is a lack of production planning, leading to overload in certain work centers. These results are significant for several reasons. Firstly, the knowledge obtained through data analysis is directly linked to the quality of the data. Therefore, an organizational culture that values such performance monitoring activities should be promoted, as organizational performance depends on them. On the other hand, identifying workload overload in work centers enables alerting production planning managers about this issue, as it affects the quality and delivery deadlines of products or services, as well as employee motivation.
In this project, the aim was to analyze and investigate organizational data in order to identify areas for improvement and adjustments. Therefore, extensive exploratory analysis was conducted, based on performance indicators. Additionally, dashboards were developed to meet user needs, taking into account aspects such as visual complexity and the relevance of the presented data. Furthermore, the implementation of a sales forecasting algorithm was studied, considering methods such as the Additive Holt-Winters method and the SARIMA method. This analysis was performed considering the stationarity evaluation of the time series data and the presence of seasonality in the data. Finally, based on these findings, some flaws were identified regarding the data input process in the organizational ERP. Moreover, it was observed that there is a lack of production planning, leading to overload in certain work centers. These results are significant for several reasons. Firstly, the knowledge obtained through data analysis is directly linked to the quality of the data. Therefore, an organizational culture that values such performance monitoring activities should be promoted, as organizational performance depends on them. On the other hand, identifying workload overload in work centers enables alerting production planning managers about this issue, as it affects the quality and delivery deadlines of products or services, as well as employee motivation.
Description
Keywords
Análise de Dados Business Intelligence Desempenho Organizacional Previsão Data Analysis Business Intelligence Organizational Performance Forecasting