Activating Latent Semantic Models with Natural Language Inference

8,120 views

This is a preview

This playground version isn't public and is work in progress.

Next Steps

To implement the base four emotional states (happiness, sadness, anger, and fear) we must first create a data structure capable of encoding the variable identifier and confidence. To keep things simple we will simply create a list containing a tuple of (identifer, confidence). This now allows us to encode mixed emotions as seen below:

fn read_emotion() -> Vec<(String,f64)> { vec![
   ("sad".to_string(),0.75),
   ("angry".to_string(),0.33),
   ("happy".to_string(),-0.95)
]}
fn draw_emotion(emotions: Vec<(String,f64)>) {
   for ref ec in &emotions {
      let ref emotion = ec.0;
      let confidence = ec.1;
      if confidence > 0.0 { println!("{}: {}", emotion, confidence) }
      else { println!("not {}: {}", emotion, confidence) }
   }
}
fn main() {
    draw_emotion( read_emotion() );
}

Open Source Your Knowledge: become a Contributor and help others learn. Create New Content

This is a preview

2/2 Mixed Latent Variables

Next Steps