Home - Abhishekkalra88/Entity-Media-Sentiment-Analyzer GitHub Wiki
Welcome to the Entity-Media-Analyzer wiki!
“Entity media Analyzer” determines a net sentiment score for the searched entity by analyzing the related web pages returned via a google search. The tool is being built to address the requirements of Anti Money Laundering Compliance department at most financial instituions. The current process involves analysts manually ("the users") researching the web pages and adjudicating a negative or positive score on the basis of their understanding of the returned web corpora. The findings are subsequently logged in a case narrative report providing the rationale for arriving at the decision taken. This entire process is largely repetitive and has to be performed for each alerted entity ( > 400K per month) and is non- standardized.
The project aims to address the same and automate the process by building a web based application (MVP) which allows for a customizable search and returns standardized sentiment scores. Further, the results would also provide information such as keywords frequently associated with the entity and help to generate the case narrative. The application will be hosted on a cloud server to allow for greater accessibility and scalability.
AWS Cloud URL : http://entity-media-sentiment-analyzer-dev.us-west-2.elasticbeanstalk.com/