Fuzzy Vocabulary Strength Algorithms
Correct me if I'm wrong, but It seems that the vocabulary strength is fixed and not adaptive or accurate to the individual. If Duolingo could replace this with a fuzzy logic system, this would be more effective for learning. The vocabulary strength should be determined by an individual's data and real time results.
If all the result data for the vocabulary are already being recorded, then it shouldn't be too difficult to write a dynamic algorithm. What do you think Duolingo staff? Feasible?
I am not sure how the fuzzy logic system works, but I think that having a dynamic vocabulary strength system would be beneficial. Since some people don't usually go back to older lessons to review, the vocabulary word strength should drop slightly based on the time between when the word was last used.
I think the words that are less used should be recycled into newer lessons so they could be reviewed or the person could just click the "Practice weakest words" button to review the words themselves. Thus, each vocabulary word is refreshed and no word is forgotten.
Fuzzy logic is a method to measure the probability of uncertain information.
In this case, the memory strength for a word is uncertain for each user ( everybody is different ). But the current system seems to be fixed; meaning it increases after a certain fixed value is surpassed; and this value is the same for everyone. This is an inaccurate representation of the actual word strength for the user.
With fuzzy logic, more information is applied in calculating the actual strength. Applying the actual individual results, would make the vocabulary strength dynamic, and also help calculate the probability of the current strength differently for each learner. Besides the result data, many other types of variables and data can be fed into this system. One uncertain variable that should be included is the rate of memory decay, which you, michisjourdi, myself and many others consider to be very important. Adding this to the fuzzy system would make the strength levels even more dynamic and accurate.