Tracking politicians & rating them based on promises fulfilled
- Get the manifesto of targeted political person / party.
- Divide the manifesto into tasks which is basic human manageble task. (eg: Supplying 24 hrs water supply to each home)
- Set up a news tracker
- Select news resources which provide news as RSS feeds or something directly consumable.
- The sources have to be a media house covering all minor to major issues happening daily.
- Decide upon the database architecture to store these news items & index them for fast skimming & searching to establish nearness to the promises.
- Have a NLP based decision system to determine whether a promise has been fulfilled by skimming through news articles.
- Decide a percentage or score to the promise based on how much it has been fulfilled.
- Rank political person / party based on scores generated.
- The task is to find the most efficient politician who walks the talk.
- Rank selected political person / party on a web based dashboard.
- Let this system continue on an automated fashion & refresh scores based on latest news articles coming.
Past period manifesto will be chosen which is old enough so that enough news articles are already accumulated. This will help in quick testing of the model & correcting wrong logic as soon as encountered.