International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 10 Issue: 08 | Aug 2023
p-ISSN: 2395-0072
www.irjet.net
Big Data Analytics for Predicting Consumer Behaviour Chandni Jivrajani1, Ghanashyam Vagale2, Ranjith koduri3, Chaithra Channegowda4, Mohammed Naif5, Matur Rohit Kumar6 ---------------------------------------------------------------------***--------------------------------------------------------------------customers' opinion of your business is consistent with Abstract - Data mining techniques are particularly their basic beliefs.
efficient tools for obtaining the hidden knowledge from a large dataset to improve predicting accuracy and efficiency. Decision analysis and predictions must be integrated into intelligent decision analytical systems. The accuracy of sales forecasts has a significant effect on business. Many corporate organizations rely heavily on their knowledge bases to forecast market trends in demand and sales. This suggested job will involve a thorough investigation and evaluation of understandable forecasting models in order to enhance future sales projections. Several data mining techniques could be used to solve these issues. In this project, the idea of sales data and sales projection is briefly examined. The several methods and metrics for sales forecasting are identified. An appropriate predictive model is given for the prognosis of the sales trend on the basis of a performance appraisal.
Customer behaviour models are built on the data mining of customer data, which is the foundation of smart budgeting. Data mining techniques are particularly effective in turning high volume of data into valuable information for cost prediction and sales forecast. Sales predictions are essential inputs for numerous decisionmaking processes at the organisational level across a variety of functional areas, including operations, marketing, sales, production, and finance. Predictive sales data is crucial for firms wanting to raise investment capital since it can be used to successfully manage internal resources inside an organisation. The studies move on from a fresh angle that concentrates on how to pick a suitable strategy to forecast sales with a high level of accuracy. The initial dataset utilised for this study included a lot of entries, but the final dataset that was used for analysis was significantly smaller than the original since it was free of useless information, duplicate entries, and irrelevant sales data.
Key Words: Big Data, Analytics, Random Forest, Linear Regression, Sales Prediction, Customer Behaviour.
1. INTRODUCTION
2. PROPOSED MODEL
In plain English, customer behaviour refers to how consumers act when shopping. It outlines the procedure customers use to decide what to buy in order to fulfil their needs and wants. It is made up of client preferences, which influence their purchasing behaviour. It is a concept that spans a number of phases, from the emergence of requirements to the choice to make a purchase. Each customer's mental state is different from the others; they are not all the same. Therefore, it is essential for every organisation to comprehend its clients. It aids businesses in satisfying client demands and preferences. Businesses that use customer relationship management should have adequate understanding of their customers. It is a database that compiles additional information about its clients.
In order to estimate how comparable consumers will behave in similar circumstances, customer behaviour forecasting is akin to developing a mathematical model to represent the typical behaviours seen among specific groups of customers. The current customer behaviour models rely on data mining methods, and each model is made to provide a single answer. For instance, this model can be used to forecast how a certain customer group will behave in response to a marketing campaign. If the model was successful, the marketer would use the same approach to draw in a growing number of clients. But because the specialists' mathematical methods and tools were so sophisticated and expensive, the current systems were more challenging and expensive. Even after developing an expensive model that was highly expensive to manipulate and process, marketers still need to know exactly what to do to attract clients to their business. However, a lot of models were overly simplistic and predicative because they omitted important features that would have complicated them. According to the aforementioned literature review, Generalised linear model, Decision tree, and Gradient boost tree were selected as a combination and used in the process. With 85,000 datasets, their best solution had an accuracy of about 64%. In the following study, data aggregation and
An examination of customer behaviour looks at both the quality and number of client interactions with your business. Buyer personas are first created by grouping customers according to their shared traits. The customer journey map's stages are then examined for each group to determine how the personas interact with your business. An analysis of customer behaviour offers insight into the various factors that affect an audience. It gives you a glimpse into the motivations, objectives, and procedures used for making decisions across the customer experience. This study enables you to determine whether the
© 2023, IRJET
|
Impact Factor value: 8.226
|
ISO 9001:2008 Certified Journal
|
Page 156