Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Extending Deep Knowledge Tracing: Inferring Interpretable Knowledge and Predicting Post-System Performance, Richard+ (w/ Ryan Baker), ICCE'20 #476

Open
AkihikoWatanabe opened this issue Aug 29, 2022 · 1 comment

Comments

@AkihikoWatanabe
Copy link
Owner

https://arxiv.org/pdf/1910.12597.pdf

@AkihikoWatanabe
Copy link
Owner Author

AkihikoWatanabe commented Aug 29, 2022

概要

ざっくりとしか読めていないが

  • DeepLearningBasedなKT手法は、latentな学習者の知識を推定しているわけではなく、「正誤」を予測しているだけであることを指摘
    • → 一方BKTはきちんとlatent knowledgeがモデリングされている
    • → 昔はknowledge inferenceした結果を、post-testで測定したスキルのmasteryとしっかり比較する文化があったが、近年のDeepLearningベースな研究では全く実施されていないことも指摘
    • → learning systemの中でどのようなパフォーマンスが発揮されるかではなく、learning systemの外でどれだけスキルが発揮できるか、というところにBKTなどの時代は強い焦点が置かれていたのだと思われる
  • DeepLearningBasedなKT手法でもknowledgeのinferenceが行える手法を提案し、BKTやPFAによるknowledge estimateよりもposttestのスコアと高い相関を示すことを実験した
    • → 手法: それぞれの問題のfirst attemptに対する正誤データの「全て」をtraining dataとし、DKT, DKVMN, BKT, PFAを学習。
       -(おそらく)学習したモデルを用いてある生徒AのスキルBのknowledgeをinferenceしたい場合、生徒Aが回答したスキルBと紐づいた問題に対する平均正答率を推定した習熟度とした
       - 生徒Aはtraining dataに含まれている生徒
    • すなわち、生徒Aにとって未知の問題の正答率を予測しているわけではなく、モデルがパラメータを推定するために利用した既知の問題-回答ペアデータに対して、モデルがパラメータをfittingした後にinferenceできる正答率の平均値を習熟度としている

結果

  • 4種類のスキルに対するpost-testのスコアと相関係数をモデルごとに比較した結果、DKT, DKVMNなどは、BKTよりも高い相関を示し、PFAとはcomparableな結果となった
    image

所感

  • この手法のリアルタイムな運用は難しいと思った(knowledgeをinferするために毎回モデルをtrainingしなおさなければならない)
  • BKTが推定するスキルのmasteryはこのcase studyだけ見ると全くあてにならない・・・
  • ユーザが回答した問題と紐づいたスキルのknowledgeしか推定できないところもlimitationの一つだと思う
  • この手法がtraining dataに含まれていない「未知の問題」に対する正答率予測を平均することで、knowledgeをinferenceできるという話だったのであれば、非常に興味深いと思った。
     - 実際どうなんだろうか?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

No branches or pull requests

1 participant