A carregar...
Publicação
Site Selection Using Geo-Social Media: A Study For Eateries In Lisbon
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 5.69 MB | Adobe PDF |
Autores
Orientador(es)
Resumo(s)
The rise in the influx of multicultural societies, studentification, and overall population growth has positively impacted the local economy of eateries in Lisbon, Portugal. However, this has also increased retail competition, especially in tourism. The overall increase in multicultural societies has also led to an increase in multiple smaller hotspots of human-urban attraction, making the concept of just one downtown in the city a little vague. These transformations of urban cities pose a big challenge for upcoming retail and eateries store owners in finding the most optimal location to set up their shops. An optimal site selection strategy should recommend new locations that can maximize the revenues of a business. Unfortunately, with dynamically changing human-urban interactions, traditional methods like relying on census data or surveys to understand neighborhoods and their impact on businesses are no more reliable or scalable. This study aims to address this gap by using geo-social data extracted from social media platforms like Twitter, Flickr, Instagram, and Google Maps, which then acts as a proxy to the real population. Seven variables are engineered at a neighborhood level using this data: business interest, age, gender, spatial competition, spatial proximity to stores, homogeneous neighborhoods, and percentage of the native population. A Random Forest based binary classification method is then used to predict whether a Point of Interest (POI) can be a part of any neighborhood n. The results show that using only these 7 variables, an F1-Score of 83% can be achieved in classifying whether a neighborhood is good for an “eateries” POI. The methodology used in this research is made to work with open data and be generic and reproducible to any city worldwide.
Descrição
Dissertation submitted in partial fulfilment of the requirements for the Degree of Master of Science in Geospatial Technologies
This thesis is one of the outputs of a funded project by the Fundação para a Ciência e a Tecnologia (FCT). The project is named as "CityMe" with the Project Reference (EXPL/GESURB/1429/2021).
This thesis is one of the outputs of a funded project by the Fundação para a Ciência e a Tecnologia (FCT). The project is named as "CityMe" with the Project Reference (EXPL/GESURB/1429/2021).
Palavras-chave
Site Selection Geographic Self-Organizing Maps Geospatial Analysis Retail Social Media