Twitter Geographic Sentiment Analysis
The EurekaFacts Sentiment Analyzer has three modules: capture, analysis, and output. When a keyword is selected, the Capture Module reaches out to a social media platform, collects available interactions (this example examines the most recent set of fifty tweets), then the Analysis Module examines the content for tone and geographic location of origin, and produces in the Output Module, a map that represents negative comments in red, positive in green, and neutral in blue. When set in a time-based visualization, reactions to different topics can be detected as well as their changes over time.
The EurekaFacts Sentiment Analyzer Capture module has been adapted to collect data from Twitter. In other applications, we can develop capture from:
- an aggregator of media and of print advertising comprising both text and images
- large data feeds of social media postings both directed and of search-based capture (we use content terms and hashtags)
- input from crowd-sourced efforts
The EurekaFacts Sentiment Analyzer Analysis module has been set to deliver a simple sentiment analysis based on the context of various posts. Other applications can be trained for more complex analyses and categorizations.
The EurekaFacts Sentiment Analyzer Output module uses geography and the content categorization. Other Output modules can be configured such as
- plug-and-play feeding into other actions
- actions prompting human intervention
- setting in motion other processes
- delivering a report
- production of summary reports (yes, we can have the AI assistant do that too!).