| الوصف: |
Through this thesis we set the hypothesis that, via the creation of a set of custom toolkits, using cloud computing, online user-generated content, can be extracted from emerging large-scale data sets, allowing the collection, analysis and visualisation of geospatial data by social scientists. By the use of a custom-built suite of software, known as the 'BigDataToolkit', we examine the need and use of cloud computing and custom workflows to open up access to existing online data as well as setting up processes to enable the collection of new data. We examine the use of the toolkit to collect large amounts of data from various online sources, such as Social Media Application Programming Interfaces (APIs) and data stores, to visualise the data collected in real-time. Through the execution of these workflows, this thesis presents an implementation of a smart collector framework to automate the collection process to significantly increase the amount of data that can be obtained from the standard API endpoints. By the use of these interconnected methods and distributed collection workflows, the final system is able to collect and visualise a larger amount of data in real time than single system data collection processes used within traditional social media analysis. Aimed at allowing researchers without a core understanding of the intricacies of computer science, this thesis provides a methodology to open up new data sources to not only academics but also wider participants, allowing the collection of user-generated geographic and textual content, en masse. A series of case studies are provided, covering applications from the single researcher collecting data through to collection via the use of televised media. These are examined in terms of the tools created and the opportunities opened, allowing real-time analysis of data, collected via the use of the developed toolkit. |
