DATA MINING FOR ANALYSING TRENDS AND CUSTOMER
                 BEHAVIOUR VIA INTERNET SEARCHES

                                               Mioara POPESCU


Abstract: The potential of analysis the web searches has been drawing attention of the scientific
community for a few years. Despite the large amount of data publicly available from the internet
searches, the opportunities for more advanced analysis are still relatively unexplored. The main
objective of this paper is to analyse how the web searches can be used as an indicator for
different marketing campaigns and strategies. The first analysis explores the possibility to
measure the impact and results of the marketing campaigns by the volume of web searches from
the period when the campaign was implemented. The second analysis aims to predict the
customer behaviour based on the web searches, from the keywords correlation. Public data
extracted from search engines has been used for both analyses together with tools used for
analysis of trends and correlations of web searches.

Keywords: Data Mining; Web Mining; Correlation; Trends; Pearson Coefficient

JEL Classification: C53, J11, E24