Page 1

IJSRD - International Journal for Scientific Research & Development| Vol. 4, Issue 02, 2016 | ISSN (online): 2321-0613

Reducing Complexity of Business Forecasting using Big Data Predictive Analysis Rahul Yadav1 Reetika Singh2 1 Research Scholar 2Faculty 1,2 Department of Computer Science & Engineering 1,2 Shri Shankaracharya College of Engineering & Technology, Chhattisgarh Swami Vivekanand Technical University, Bhilai - 490006, Chhattisgarh, INDIA Abstract— The Big data presents most of its consequence from the perceptions it produces when analyzed finding patterns, deriving text meaning, creating conclusions and finally responding to the world with intelligence. Businesses are turning towards the predictive analytics to help them build the engagement with customers, optimize processes and reduce operational costs. Predictive analysis is usually used to forecast future probabilities which can arise in future for business. We can make the business strategy for the predictive models used for the analysis of the current data and historical data in order to build up a good customer’s relationship and relevant products of company and to identify potential risks and also opportunities for a company. It uses a number of methodologies, including data mining, mathematical modeling and also machine learning to help the analysis process more easier for future business forecasts. In this paper work, we are presenting an approach to implement Business forecasting application which can be used as user friendly tool for the Predictive analysis. In our approach we are simplifying the complexities involved using Standard Advanced Predictive Analysis Tools which requires advanced skills and expertise to use in Business Forecasting. Key words: Big Data, Predictive Analysis, Business Forecasting, Visualization Tools, Data Mining

can easily filter it out or we can provide a data cleansing for getting analysis of the data. Predictive analytics is the use of past/historical data to predict future trends. The predictive model makes the implementation of the statistical models and machine learning algorithms for the identification of the patterns and learning from the historical data. The advent of big data analytics will be more informed about the data driven strategies for businesses communicate with consumer. Such that different vendors of the online shopping portal and the online business company can analyze the tremendous amount of the data generated from Electronic Data Interchange (EDI) for the better understanding of consumer behavior. Such unique insight can be applied to increase customer service, guide business strategy and provide user friendly services to customers. The paper is described below in different section as follow. In Section 2 of the paper we will discuss about the related work for predictive analysis of the sales records and similarly in Section 3, we will propose the system architecture and some data prediction with the filtering algorithms for the analysis. In Section 4, we present real experimental result for different stores based on historical data. At last, we conclude the paper in Section 5.

I. INTRODUCTION

II. RELATED WORK

In present world the development of business and economic growth has raised the amount of data usage over the internet and also it has being over exploited with its size. As a result, Big Data have become a new buzzword in information technology. Collecting, managing, and analyzing Big Data is a challenge and will soon become a major companion between high and low performing organizations. This section deliberates the issue of Big Data including the four proportions of Big Data and the opportunities along with the challenges created by them. It also discusses about the Big Data analytics applications. Every day, we use many devices to create large amounts of data for example, searching online, visiting of online shopping portal for making purchases of the products and also making transactions in the shopping portals, reading data from sensors, using social media platform to connect with our friends also the usage of the location based system which points our location using the GPS services. All the data are collected and stored in several databases technology which forms as the Big Data. In this research paper we will analyze about the sales and increase in number of production of goods in stores and what are the key factors to build up the better customer’s relationship. When the store owners want to filter out the related data from their dataset of past and historical data they

 Introduction to Predictive Analysis From last few years the predictive analysis has become a great visualization tool for forecasting and analysis of business strategies for the growth of the business. Analyze the existing data stores to make a selection of the predictions, such as where you might need more storage or how to make customers attract towards your store and increase the sales and revenue. Many tools assured to make predictions but, in truth tend to do more of a historical and past analysis of data at present time. In Uyoyo Zino Edosio proposed a paper that as much as big data analytics hold many promises for providing business perceptions and analyzing the consumer behavior also with its unique challenges. According to the research, largest obstacles to big data analytics are the staffing and training of the big data user, which are followed by privacy constraints. Bulk of the consumers are concerned about how their personal confidential information about the customers is kept in the database with secure encryption of the data. Despite these challenges a lot of companies are moving ahead to implement big data in their e-commerce strategies.

All rights reserved by www.ijsrd.com

387


Reducing Complexity of Business Forecasting using Big Data Predictive Analysis (IJSRD/Vol. 4/Issue 02/2016/114)

Fig. 1: Process of Predictive Analysis The paper intention is to validate that big data analytics can be used as a foundation for the increasing profit of the organizations by improving the business parameters. Furthermore, we will utilize the case studies and also the paper aims about how to establish that big data analytics supports conception, enhancement of various business services to expressively improve the customer understanding as well as worth making for organizations. The present research work deals with the detailed description of big data’s technology predictive analysis working and calculation of forecasting issues faced by the enterprises on their way to adopt big data technology for contentment of their business need and for making some relevant decision. An efficient and integrated solution for big data predictive analysis is proposed by performing comparative analysis of existing solutions for big data challenges and also forecasting of the business growth and revenue in terms of business. The effective approach to avail business forecasting services is to NO-SQL model for scrabbling infrastructure as per business requirements by providing the required data’s to business Owner. III. PROPOSED WORK The proposed solution will mainly focus on predictive analysis and also the data mining of the required data from the collection. The user will have their stores and past/historical data of there stores from previous transactions which will provide user with more predictive way that how to increase the sales and production of products. There are two main algorithms/technologies that will support the recommendation system in data mining: Collaborative filtering:One approach for making the collection of the stores and products recommendation bassed on the past revenues and historical data.Collaborative filtering methods are based on collection of huge datasets and analyzing a large clusters of information on the users behaviors, activities or preferences and predicting what the users will suggest and like according to their usage of the product as compare to the other users. A key advantage of the collaborative filtering approach is that it does not depend upon the machine learning contents and therefore it is capable of accurately recommending complex items such as movies, videos without requiring an "understanding" of the data. Many algorithms had been used in the measuring of user or item comparison in recommended systems.

Fig. 2: Collaborative Filtering Algorithm Cluster analysis or the clustering is task of collecting a set of objects in such a way that objects in the same group called a cluster are more analogous (in some other sense or another way) to each other than to those in other groups (clusters). The task of performing the exploratory data mining and common technique for statistical data analysis, used in many fields of the machine learning, pattern recognition, information retrieval and bioinformatics. In this paper we several algorithms are designed for this task, including techniques that are based on correlation coefficients, vector-based matrix calculations and statistical Bayesian methods. We compare the predictive precision of the various methods in a set of small stores and all the sales of different stores. We use two basic classes of evaluation metrics. The first feature will be the accuracy over a set of individual predictions in terms of average absolute deviation. The second important characterizes is the utility of a sorted list of profitable stores. This metric uses an approximation of the probability that customer will see a prediction chart in the form of comparison chart on the sales forecasting. IV. RESULTS In result, the usage of the different data sets and collection of the many stores,products,online portal for analysis of the data for better result.Using the Big Data which is collected are minified using the filters and to develop the high performance concurent program that don't rely on the mainstream of the multithreading approach but use asynchronous I/O with an event-driven programming model such as Node.js which is gaining popularity in the server-side JavaScript space. Online Shop vendors can efficiently predict consumer behavior faster, more proficiently and at more effective cost in production.

Fig. 3: Backend project structure using Node.js (MEAN Stack)

All rights reserved by www.ijsrd.com

388


Reducing Complexity of Business Forecasting using Big Data Predictive Analysis (IJSRD/Vol. 4/Issue 02/2016/114)

learning algorithm), Online Shop vendors can efficiently predict consumer behavior faster, more proficiently at more effective cost in production. VI. CONCLUSION

Fig. 4: Mongo DB Service for storing training samples as Mongo Collections to Simulate Big data source running on default port 27017

In this research work we are presenting an approach to implement Business forecasting application which can be used as easy to use Predictive analysis tool using Node.js in the backend along with MongoDB as our data source which will be of less memory space and will perform more efficient and fast as compare to the other data objects. In our approach we are simplifying the complexities involved using Standard Advanced Predictive Analysis Tools which requires advanced skills and expertise to use in Business Forecasting. The main objective of this research is to provide the predictive analysis of the historical data and suggesting the new business strategies also providing the user friendly data presentation using some graphs and solving all the problems faced by the different online vendors. REFERENCES

Fig. 5: Graph from recent survey made by Online Vendors

Fig. 6: Showing the sales prediction of different Stores. V. DISCUSSIONS In summary, we can say that predictive analytics based on the definitions above of the selection techniques, deals with use of data to define and identify possible future events. The knowledge has been around for a while, though the implementation of the system is low because of the complexity and cost. Using the Big data analytical platform to analyse these data (along with the data mining and machine

[1] Yaakub, Algarni M.R., A. Bo Peng. (2012).Integration of Opinion into Customer Analysis Model, in Web Intelligence and Intelligent Agent Technology (WI-IAT) IEEE/WIC/ACM International Conferences on, Vol.3.14-16 [2] Balar, A. Malviya and N. Prasad, S. Gangurde, A.(2013).Forecasting consumer behavior with innovative value proposition for organizations using big data analytics in Computational Intelligence and Computing Research (ICCIC), IEEE International Conference on , Vol.2. pp.1-4 [3] Chi-Hwan Choi.(2013).Voice of Customer Analysis for Internet Shopping Malls in International Journal of Smart Home Vol.7. No.5, pp.291-304 [4] Fung, H.P. (2013). Using Big Data Analytics in Information Technology (IT) Service Delivery. Internet Technologies and Applications Research, Vol.1 (1), pp.6-10. [5] Mosavi A., Vaezipour A. (2013). Developing Effective Tools for Predictive Analytics and Informed Decisions. Technical Report. University of Tallinn. [6] Seave, A. (2013). Forbes.com: How Social Media Drives Your Customers' Purchasing Decisions. (Online) Available at: http://www.forbes.com/sites/avaseave/2013/07/22/howsocial-media-moves-consumers-from-sharingtopurchase (Accessed 29 Jan 2016) [7] Wielki J. (2013). Implementation of the Big Data concept in organizations – possibilities, impediments and challenge, In FEDCSIS. 2013. [8] Krumeich J., Jacobi, S., Werth D., Loos, P.(2014).Big Data Analytics for Predictive Manufacturing Control - A Case Study from Process Industry, in Big Data (Big Data Congress), 2014 IEEE International Congress on ,Vol.1(1). pp.530-537, June 27 2014-July 2 2014 [9] Greene & William. Predictive Analysis. https://en.wikipedia.org/wiki/Predictive_analytics. (Accessed 29 Jan 2016). [10] Prasad Babu, M.S. Sastry and S.H.(2014).Big data and predictive analytics in ERP systems for automating

All rights reserved by www.ijsrd.com

389


Reducing Complexity of Business Forecasting using Big Data Predictive Analysis (IJSRD/Vol. 4/Issue 02/2016/114)

decision making process in Software Engineering and Service Science (ICSESS), 5th IEEE International Conference on , Vol. 2. pp.259-262, 27-29 June 2014 [11] Uyoyo Zino Edosio. (2014).Big Data Analytics and its Application in E-Commerce. , International Conference of E-Commerce Technologies, At University of Bradford, 1(1).24-27

All rights reserved by www.ijsrd.com

390

Reducing Complexity of Business Forecasting Using Big Data Predictive Analysis  

The Big data presents most of its consequence from the perceptions it produces when analyzed finding patterns, deriving text meaning, creati...

Reducing Complexity of Business Forecasting Using Big Data Predictive Analysis  

The Big data presents most of its consequence from the perceptions it produces when analyzed finding patterns, deriving text meaning, creati...

Advertisement