Pré-processamento e Análise de Dados de Táxis

• Cristiano Martins Monteiro
• Fábio Rocha da Silva
• Cristina Duarte Murta

DOI: https://doi.org/10.5753/semish.2017.3368

Resumo

The study of large amounts of data is a current challenge and we must be prepared to treat and analyze them. In this task, pre-processing is essential for verifying data, identifying inconsistencies, possible errors and incompleteness. In this work, two datasets with more than thirty million records of the movement of taxis in the cities of San Francisco and Rome were analyzed. We propose an algorithm to treat anomalous speeds identified in the preprocessing step of these datasets. We present the analysis of the datasets before and after the application of the algorithm, showing its relevance and pertinence. The results show specific characteristics of the taxi service in the two metropolises.