View Materialization Over Big Data

2 The Big Data Architecture For Large Data Stores

Big data is not just the data of multiple thousand entities but the repetitive data generated by such entities that make data Big (Jacobs, 2009). For example, the number of users, products, vendors etc. of e-commerce web sites may not make the data Big, but storing the number of transactions and other actions made by these users over the e-commerce application is what that makes the data Big. (Jacobs, 2009) suggested that a database of size more than 100 GB involving joins on non-key attributes, will require potentially very large computing resources and, therefore, cannot be considered as small data. Big data is also heterogeneous, as it includes structured data (Relational data); semi-structured data (XML or similar object data); unstructured data (text, voice, audio), data from web (social media, blogs, web logs, click streams); spatial data (coordinates, GPS data); data from sensors and RFID; and scientific data.