Twitter Data Mining to Map Pedestrian Experience of Open Spaces

Abstract

This research investigated the classification and visualisation of Twitter user-generated data. Twitter data were classified based on their sentiment relating to pedestrian experience of the quality of open spaces, based on their content. The research methodology for Twitter data collection, processing and analysis included five phases: data collection, data pre-processing, data classification, data visualisation and data analysis. The territorial focus was on Oxford Street, London, UK. Special attention was placed on the questions regarding the potential of using Twitter data for extracting relevant topics for the public space and investigating whether the sentiment for these topics can relate to urban design and improvement of pedestrian space. The proposed research model considered amount and relevance, its possibilities regarding the interpretation of the collected sample, the potential of the data for the purpose of the analysis of pedestrian space quality, the precision of sentiment determination and the usability of data in relation to a particular open public space.