ePubs

The open archive for STFC research publications

Full Record Details

Persistent URL http://purl.org/net/epubs/work/34384245
Record Status Checked
Record Id 34384245
Title Scientific Database Support for Materials Modelling
Contributors
Abstract Scientific workflow is a stream of jobs, which have to be performed in an ordered manner for any scientific experiment. These workflows are often very complicated and need different sources, such as secured internet services or high performance computing services, where from data have been procured and used. The current advancement of computing power encourages scientists to do more and more computational simulation experiments, which generate a huge volume of data. Materials Modelling is a branch of science rapidly developing in recent years. scientists in this field are not capable of handing the volume of data generated, which is often few hundreds of Gigabytes to Terabytes, and require support from computer scientists to manage their workflow and data. These data should be stored in some useful manner so that scientists can use it in future to extract useful information for their research. Although there is a number of scientific workflow management software available, they are dedicated for a particular kind of experiment and no such software is available for Materials Modelling. These software don’t provide the requirements of data storage as well. In this project a software has been developed to provide a scientific workflow and database support to experiments of Materials Modelling simulation work. A generic approach has been taken so that this software will be useful for any this kind of scientific workflow. To overcome the complication of several data formats used by materials modelling scientists a consistent data format XML has been used. Wrappers have been developed to wrap different kinds of data format. Since this is the first approach of this kind of work, XML schema and DTD have been developed and a vocabulary has been developed in collaboration with the scientists in this field. The Berkeley DB XML has been used for database support at three different points of the workflow, however, the software has been kept generic for including any this kind of database managing software. A querying facility has been developed on three databases using XPath and XQuery. A user interface GUI developed also to provide a friendly platform to scientists who are not familiar with XML technology or databases. For testing the project software experiments have been done on three databases having different volume of data. The experiments show that for simple XPath type query it will be good approach to submit a number of queries at a time, because computation time required for opening and closing databases are 80% more than to execute a simple query. For complex query, however, a single query at a time is useful. It has been indicated that in future optimisation of DTD and query have to be done simultaneously for increasing efficiency of this software.
Organisation ISIS , STFC
Keywords
Funding Information
Related Research Object(s):
Licence Information:
Language English (EN)
Type Details URI(s) Local file(s) Year
Thesis MSc (Computer Science) Thesis, University of Oxford, 2011. 2011