Literature Review

       In a research on spatial analysis of Social Media Content (Tweets) during the 2012 U.S. Republican Presidential Primaries, (Tsou and Yang 2012) have used the spatial-defined query function provided by Twitter Programming Interface (API) to compare popularity of different parties by using the full name of candidates as search keywords. Search results were limited to the target area through city boundaries. R scripts were then used to calculate the correlation between different tweet variable e.g. tweets two days before elections, tweets one day before election etc. As a part of conclusion, it was noted that the analysis of tweets content might enable the researchers to identify relations between different event and spatio-temporal human behavior.

       Mapping of social activities like Twitter have been used by a number of researchers to Perform required analysis. (Malleson and Andreson 2013) used social media data to assess spatial crime hotspots. In this research identifies the areas prone to high crime rates using residential census data and ambient crowd-sources population. Two statistical methods were used to map hotspots for violent crime rates and the results were compared. The results indicate that the population data sets and statistical analysis method used effect the hotspot results except few neighborhoods.

       Another hot research area is of monitoring disaster impacts, now the question is of how to use the geo-tagged social media information to create a real-time awareness system about disaster. This is addressed by (Jung, Tsou, and Issa 2015), in this research a web application named GeoViewer was made using open source libraries and databases. The researchers have targeted tweets about wildfire and have made the web-based interface using jQuery and Leaflet APIs to interact with the tweet records stored in backend MongoDB. The filtration process was done using python scripts. Other than the web component, the spatiotemporal analysis of evacuation tweets during wildfire events were also analyzed and represented using spider chart and line graphs.

       As far as geo-related research on social media platforms like twitter is concerned, it is notable that a lot of work has been done. Addressing elections particularly, GIS has been used mostly used at the post-elections phase e.g. to visualize the results of election district wise, but has not been used extensively in the pre-election phase e.g. like our project which focuses on identifying the campaign sites for efficient uses of resources like time and cost. We can compare our project with the previous related work and solution in terms of need of research, tools and techniques used, end-user benefit etc.