Pattern based partitioning and distribution for sensor data

Abstract

"Nowadays there has been a rapid growth in the number of RDF data based applications. The data requests are of categories like real time request, critical contents, swift analysis etc. Query Processing Time is a very big factor in deciding the performance for such applications based over the triple structure of RDF i.e. subject, property and object. These application’s data-sets are very vast in terms of size and needs a redesign. The distributed architecture helps in utilizing the memory as well as storage limits in an efficient manner. The processing power available from different computing machines in the distributed architecture helps in solving such mentioned requests simultaneously without adding load only to a single sever. We will be utilizing a distributed architecture to translate the given RDF dataset into a relational schema to enhance query processing. The findings in the thesis will help in accelerating the query processing time contributing towards faster speed of forecasting of weather with the aid of sensor data. It will help in avoidance and prevention of calamities and also aid to the improvement in weather forecast methods."