Sequence Mining Analysis on Shopping Data

• Autor(a) principal: João Miguel da Rocha Ribeiro
• Data de Publicação: 2017
• Tipo de documento: Dissertação
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: https://repositorio-aberto.up.pt/handle/10216/103492
• topic: Engenharia electrotécnica, electrónica e informática; Electrical engineering, Electronic engineering, Information engineering

Being so easy to have access to information it's only natural that people and companies try to extract the maximum real value from it. Every large retail stores and commercial centres in the world fight to have the opportunity to be in possession of data about their customers and habits. This data has been extracted through the use of data mining techniques. Due to this relentless demand for new data, every new mean of finding it can bring great competitive advantages over other competitors.This dissertation presents a group of analyses made to a dataset composed by stores' visits. There are already several types of tests made to datasets of this kind in order to better understand the clients. However, the sequence mining techniques are rarely used. These techniques' main goal is to analyse a large set of data with a sequence temporal format and extract the set of sequences with similarities between all the elements. By applying these techniques correctly in a sequence dataset we can find that they can help to extract different and quality information.The dataset is composed of real spacial-time data from clients' locations in a commercial centre. Each element of this data contains a client ID, a store, the specific time of that detection and other information. Through these elements, different types of sequences can be made. The dissertation presents some of these possible sequences as well as the types of sequence mining analyses performed on each one.