Competitive Edge Using Big Data Analytics to Improve Customer Relationship Management

In order to handle e-commerce data for customer behaviour analysis, bigdata analytics is a developing technology that shows promise. The ability to analyse heterogeneous data in native format and at a comparatively low cost is one of big data's advantages. Data is being produced in various structures, or even unstructured, due to the rise of social media and personal digital assistants. The amount of processing time can be significantly reduced by using big data technologies, even though the data is enormous, by performing parallel processing. Metrics like consumer loyalty, affinity, transaction value, and likelihood of purchase can all be projected by machine learning algorithms. In addition to helping the shop adjust business strategy, this also helps them add inventory and run marketing. In this thesis study, two models are proposed: the Enhanced Model for Customer Behavior and Purchase Analysis and the Mouse Movement Pattern based Analysis of Customer Behavior (CBA-MMP). Two methods are used to classify the customer: the Decision Tree algorithm and Multi-Layer Neural Networks (MLNN). Decision trees employ metrics including gender, age, day segment, special occasion, total time spent on-site, and purchasing done or not done. Additionally, lowering processing times and increasing accuracy are the goals. Processing the input with a collection of hidden layers and an output layer is done using a multi-layer neural network. Every layer node has an activation function attached to it. Based on a comparison investigation, the suggested model yields a high accuracy rate with a reduced processing time.