Enrichment patterns for big data

Abstract

Importance of 'Big Data' in terms of business value is very well understood across different sectors such as telecom, banking, insurance etc for targeted campaigns or real time performance actions. 'Big Data' emphasizes the following characteristics, Velocity, Volume, Variety, and Veracity. Business adopts one or more of the above properties to cater to the requirements of the clients. Data being crucial in this case has different facets. The sources of data being different and consumption across different businesses makes the data modeling a tougher problem. Data schema evolves with new sources of data, changes due to change in data sources, etc Thus enrichment of data constantly triggers the needs to device methods to adopt the models to the new patterns. When the enrichment patterns are understood, modeling the Big Data and Management becomes easy. We highlight the list of such identified patterns based upon our real world implementations. In this work, we propose a method to evolve the data models from its initially defined schema such that data models can easily adapt to changes. We show through cases studies from real world example that our model can adopt to evolve data from different sources.