Skip to content
Jaysin edited this page Feb 29, 2016 · 14 revisions

Future Plans

  • Build a larger message database
  • Use sudo Bayes Theorem by giving values (weights) to each beginning message and ending message and decreasing their values (weights) for a certain person that has been sent that particular message.
  • For example, Let's say all messages start off with a base rating of 10. If I send Amy Beginning: "Hey Amy just wanted to tell you" Ending: " I miss you" both that beginning and ending are decremented by one for Amy's rating. Now, B_a += 1 = 9 and E_a += 1 = 9. If we always choose beginning and ending that is higher than the highest message rating (100 <-- 10*10) then all combinations of messages will be executed before starting to repeat any messages.
  • This way, not only will the message seem more authentic and believable but the limitations will always be known based on the size of the beginning and ending message sets for that particular person.
  • Provide stand alone desktop applications for Mac, PC, Linux.
  • Mobile application.
  • Who knows what else?
Clone this wiki locally