Use this URL to cite or link to this record in EThOS:
Title: Shopping behaviour forecasts : experiments based on a fuzzy learning technique in the Spanish food retailing industry
Author: Casabayó Bonàs, Mònica
Awarding Body: University of Edinburgh
Current Institution: University of Edinburgh
Date of Award: 2005
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
The general aim of this thesis is to analyse the possibility of developing synergies when connecting 3 different areas of research namely consumer behaviour, market research and artificial intelligence (AI). The three areas of research are very extensive. When analysing the potential links between them, a wide number of triple combinations arise. In addition, the number of combinations can also be increased when applied to different industries but the food retailing industry is selected as the framework of this thesis. A general overview of the three disciplines is developed. Firstly, consumer behaviour fundamentals are interpreted and reconsidered from a food retailer’s perspective. This constitutes one approach to the research in this thesis. Secondary, considering that learning from past data to anticipate shopping behaviours is a retailer’s focus of research, an overview of the main market research forecasting models and techniques is carried out. Thirdly, machine learning (AI subfield) is also explained in respect of its capability to perform forecasting tasks. Handling customer data is not easy. Information tends to be ambiguous, uncertain and incomplete. Moreover, the customer behaves differently according to his/her situation. Another AI subfield, fuzzy logic (Zadeh 1965) is also explained as it copes with the concept of partial truth. Having reviewed the three disciplines, the triple combination of ‘shoppers (household)’, ‘forecasting behaviours’ and ‘fuzzy learning’ aspects from each mentioned domain respectively are selected as illustrates the scope of this thesis. The empirical research consists of two experiments focused on forecasting shopper’s behaviour (in particular household shopping behaviour), in the food retailing industry using LAMDA (a fuzzy learning technique).The methodology of research is mainly based on data extracted from a Spanish Food Retailer’s (Supermarcats Pujol SA) databases. The first experiment is based on LAMDA’s supervised learning approach and provides a model to forecast the current customers who are going to defect when a competitor opens a supermarket in the same area. The second experiment is based on LAMDA’s unsupervised learning approach and provides a model to forecast the current customers who are going to buy online once the company launches the Website. Results indicate that marketing expert’s judgements are a key point when using learning techniques to forecast behaviours. Customers are not simple robots. People may change their behaviour according to their situation. The results show that when applying the adequacy degree (fuzzy logic concept), the forecasting accuracy increases considerably.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available