A best source would be twitter to grab news events. Given that our model is a daytrader model <=10min execution... we would need a more fast and consistente source of new information.
** Potential issue on this is that we are currently using ibmwatson to parse the news score, if we increase the feeds we may hit the api limit calls i.e. we would need to find a new way to generate news Natural Language Processing scores.
solution here done by @mgilliland23 was to remove the current google news api integration with ibm watson and move to twitter api integrating with ibm watson - that seems stable for now so I'm closing the issue.
A best source would be twitter to grab news events. Given that our model is a daytrader model <=10min execution... we would need a more fast and consistente source of new information.
** Potential issue on this is that we are currently using ibmwatson to parse the news score, if we increase the feeds we may hit the api limit calls i.e. we would need to find a new way to generate news Natural Language Processing scores.