Analysing Customer Buying Habits with Visual Data Mining

Abstract

The ultimate aim of every industry and organization is to make profits by attracting a greater number of customers. To to achieve this motto they need to analyze the priorities of the customer. This is usually done by doing marketing everywhere such as in social media, newspaper, sites, etc. The marketers keep advertising at many sites without even knowing whether the advertisement is useful at that platform or not. Hence, making such huge investments in advertisements at wrong platforms will lead to less profits. There is a need for the companies to provide customer services in such a way that the customer doesn’t lose his interest and trust and must maintain a healthy relationship. If such services are provided equally to every customer, then there is a possibility that the company might provide its service to those customers who bring low profits and keep the customers who make high profits in waiting. To avoid such discrepancies, categorization of customers can be done based on their priority. This can be achieved using the Clustering technique, k-means algorithm. Since the customer data is unsupervised, k-means helps us to cluster them. If we use supervised data, then prediction of new customer’s priority can also be done using K nearest neighbors’ algorithm. The exploration of deep insights of data using exploratory data analysis makes it easy to understand data using visual representations [3][4][6]. These visual representations also lead to less time consumption for exploratory data analysis.