EXISTING OPERATIONAL TECHNIQUES TO STORE/PROCESS BIGDATA

The role of cloud computing towards educational excellence has been discussed by various prominent authors e.g. (Wang & Xing, 2011; Sqalli et al., 2012), most recently. Various prominent educational institutions around the world have started with adoption for public cloud and at present some of them are also moving to the private cloud. 2-3 years back the educational data uploaded on clouds were usually course materials or some other resources with less complex problems pertaining to type of data. But at present, even the educational data have become much complex and posses all the 5V characteristics of Big Data (Holley et al., 2014). The high usage of ETL scripts as well as presence of numerous layers of integrations does not posses common semantics that increases the problems of data integration to manifold. Usages of ETL scripts are only found theoretically sound but in reality it is an expensive affair. There are various literatures in the existing system where researchers have identified the potential advantage of using advanced datamining technique over educational data. The study conducted by Ali (2013) have discussed that datamining techniques can be used for prediction of enrollment for prospect student, forecasting profiling of students, developing curriculum, performance of students, and many more. Study of data mining pertaining to educational data from online learning tool was seen in the work of Fernandez et al. (2014) and Thakar et al. (2015). The author have discussed about the challenges and benefits of online learning in cloud environment. Papamitsiou and Economides (2014) have adopted qualitative method to study the effectiveness of prior educational datamining tech?nique. Another researcher Ramasubramanium et al. (2009) have presented a model for datamining using set theory considering educational data. The authors used SQL and normal query language.

BigData is the advancement of conventional datamining technique. The author e.g. Daniel (2014) have investigated the impact of BigData for analyzing educational data. The study found that there are positive opportunity of applying BigData approach over i) administrator (like allocation of resources, support system, academic programming), ii) student (planning of learning, feedback system, assessment system), and lectures (enhance teaching, assist student in academic distress). Similar direction of review-based study was also carried out by Drigas and Leliopoulos (2014) and Prakash et al. (2014). The study conducted by Johnson (2014) have discussed about the standard ethics that should be considered while analyzing datamining process over educational big data.

MacHardy and Pardos (2015) have adopted Bayesian Knowledge Tracing for extracting the knowledge from educational video contents. Bayesian Knowledge Tracing is basically an approach of assessing effectiveness of knowledge mastered by instructor. The study outcome was assessed using data from various online videos with respect to mean Root Mean Square Error and significance in probability. The presented model was used for understanding behaviour of student.

In order to design an efficient operational management system for storing and processing educational BigData, it is highly essential that both the semi-structured as well as unstructured data from multiple formats and sources be emphasized. As the operational management of conventional education system highly differs from the existing educational system, therefore, development and management of any novel BigData projects toward ensuring an optimal management will required highly professional and skilled developers. It is highly important that system for operational techniques for educational data management should be user-friendly, which is not the case in the existing system (Zhao, 2015). Adoption of Big Data approach on educational system can highly enrich the quality of knowledge dissemination process and furnish a competitive learning culture. One of the core motives of any higher technical (or non-technical) education system is to use the multiple heterogeneous disciplinary knowledge to enhance the educational contents and knowledge sharing system that surfaces a pleasant environment incorporating higher dimensionality of competence and analytical skills among the learners. There are various forms of underlying and hidden patterns which are quite difficult to be explored in complex concept e.g. education; however, adoption of BigData can bring a better form of integration and associate various forms of conventional and novel data sources from multiple source points of data.

 
Source
< Prev   CONTENTS   Source   Next >