Personalized Online Analytical Processing in Big Data Context Using User Profile and Search Context

Motivation

Today, OLAP analysis from big data is entering a new phase called personalized data. Modern Big Data systems, such as Hadoop are used as an alternative for data warehousing (Hollingsworth, 2012), (Kuldeep et al., 2014). It collects and stores inherently complex data streams due to volume, velocity, value, variety, variability and veracity (Russom, 2011). According to these reasons, the traditional OLAP implementation, namely, the RDBMS based ROLAP system, seems to be inadequate, because the new massive data architectures and analytics tools go beyond SQL Datawarehouses and OLAP engines (Jie et al., 2015). On the other hand, the work (Cuzzocrea et al., 2013) highlights the raised problems and research tendencies in the field of data warehousing and OLAP analysis on Big Data. On the basis of this observation, we seek through this study to design a new architecture to solve problems when calculating OLAP data cubes.