Coleta de preços on-line de forma automática e estimação do custo da cesta básica: uma aplicação para o município de João Pessoa

• Ano de defesa: 2022
• Autor(a) principal: Felix, João Anderson da Silva
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal da Paraíba; Brasil; Economia; Programa de Pós-Graduação em Economia; UFPB
• Palavras-chave em Português: Web-scraping; Cesta-básica; Séries-temporais; Previsão; Basic-food-parcel; Time-series; Forecasting; CNPQ::CIENCIAS SOCIAIS APLICADAS::ECONOMIA
• Link de acesso: https://repositorio.ufpb.br/jspui/handle/123456789/36788

The collection of information automatically through data available on the internet has been continuously and rapidly growing for a wide range of products and services, including the collection of prices of food products. This availability of prices on on-line products opens the possibility to monitor the price variation of these products and automatically check the inflation behavior of a given region, through technologies such as web-scraping, generating a reduction in research costs and obtaining more fast de data. Thus, to take advantage of the greater speed and economy of resources that on-line data collection can provide, an algorithm was created, which automatically collects data on prices related to a set of items from a certain basic food parcel for the metropolitan region of João Pessoa. In addition, a set of four algorithms was applied to predict the price of this basic food parcel. The results indicate that the cost of the basic food parcel, estimated with prices collected in an on-line way, satisfactorily followed the estimates made by the official statistical agency of reference. Thus, the findings show that data collected online can be a reliable alternative for tracking product prices, and can be used to track inflation in a cheaper and faster way, compared to traditional collection methods.