Class MemoryModels

java.lang.Object
org.episteme.social.psychology.MemoryModels
public final class MemoryModels extends Object
Models human memory retention, recall probability, and forgetting curves. Includes implementations of Ebbinghaus's forgetting curve and spacing effect algorithms.
Since:
1.0
Version:
1.1
Author:
Silvere Martin-Michiellot, Gemini AI (Google DeepMind)

  • Method Summary

    Modifier and Type
    Method
    Description
    static Real
    ebbinghausRetention(double t, double s)
    Calculates retention probability using the Ebbinghaus Forgetting Curve.
    static double
    recallProbability(int position, int totalItems)
    Models the Serial Position Effect (Primacy and Recency).
    static double
    sm2NewEaseFactor(double oldEf, int quality)
    Adjusts the Ease Factor (EF) based on recall quality.
    static double
    sm2NextInterval(int iteration, double prevInterval, double easeFactor, int quality)
    Calculates the next optimal review interval using the SuperMemo 2 (SM-2) algorithm.

    Methods inherited from class Object

    clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

  • Method Details

    • ebbinghausRetention

      public static Real ebbinghausRetention(double t, double s)
      Calculates retention probability using the Ebbinghaus Forgetting Curve. Formula: R = exp(-t/S)
      Parameters:
      t - Time since learning (days)
      s - Stability of memory (strength), where higher values mean slower forgetting
      Returns:
      Retention probability (0.0 to 1.0)

    • sm2NextInterval

      public static double sm2NextInterval(int iteration, double prevInterval, double easeFactor, int quality)
      Calculates the next optimal review interval using the SuperMemo 2 (SM-2) algorithm. I(n) = I(n-1) * EF
      Parameters:
      iteration - The repetition number (1, 2, 3...)
      prevInterval - Previous interval in days
      easeFactor - The Ease Factor (EF), initially usually 2.5
      quality - User grade of recall quality (0-5, where >=3 is success)
      Returns:
      The next interval in days

    • sm2NewEaseFactor

      public static double sm2NewEaseFactor(double oldEf, int quality)
      Adjusts the Ease Factor (EF) based on recall quality. Implementation of SM-2 EF adjustment. NewEF = EF + (0.1 - (5-q)*(0.08 + (5-q)*0.02)) The new EF cannot drop below 1.3.
      Parameters:
      oldEf - The previous Ease Factor
      quality - User grade of recall quality (0-5)
      Returns:
      The new Ease Factor

    • recallProbability

      public static double recallProbability(int position, int totalItems)
      Models the Serial Position Effect (Primacy and Recency). Calculates recall probability based on item position in a list. Approximates the U-shaped curve where first and last items are recalled best.
      Parameters:
      position - Zero-indexed position of the item
      totalItems - Total number of items in the list
      Returns:
      Estimated probability of recall (0.0 to 1.0)