search

bigdata-ustc / EduKTM

The Model Zoo of Knowledge Tracing Models
Apache License 2.0
197 stars 63 forks source link

Question about Figure 1, 3 on the LPKT paper #29

Closed kwonmha closed 2 years ago

kwonmha commented 2 years ago

Hello, Thanks for opening codes for this project. I want to check the proficiency change of every knowledge during learning process as in the figure 1, 3. of your LPKT paper. How did you get the proficiency level from 0.0 to 1.0?

shshen-closer commented 2 years ago

Thanks for your attention. In LPKT (figure 3), for a specific KC, we have its proficiency vector in the knowledge matrix, just sum the proficiency vector and output a proficiency value, then use a logistic function to project it from 0.0 to 1.0. Here you need to try a proper logistic function, otherwise the output value maybe out of range. In figure 1, it is easier, just find the corresponding KC value in the hidden vector for students' knowledge state.

kwonmha commented 2 years ago

TY. I got it. So, the row vector of h matrix would be the proficiency vector of particular KC. And is sigmoid would be proper to get output from summed vector value?

And, for DKT, is hidden vector size of DKT same as the number of KC? For example, if I use DKT for ASSIST2012 which has 265 conceps, the hidden size should be same?