home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.ai.neural-nets
- Path: sparky!uunet!cs.utexas.edu!sun-barr!sh.wide!wnoc-tyo-news!news.u-tokyo.ac.jp!yayoi!yayoi.cc!dave
- From: dave@pipi.iis.u-tokyo.ac.jp (David Wuertele)
- Subject: Question: Has this been done?
- Message-ID: <DAVE.92Dec24160430@pipi.iis.u-tokyo.ac.jp>
- Sender: news@cc.u-tokyo.ac.jp (USENET News System)
- Nntp-Posting-Host: pipi.iis.u-tokyo.ac.jp
- Organization: Institute of Industrial Science, University of tokyo.
- Distribution: comp
- Date: Thu, 24 Dec 1992 07:04:30 GMT
- Lines: 34
-
- I would like to try to apply an ANN to the evaluation of student language
- tests. The system would have a net for each question, and the question
- i/o would be structured as follows:
- +----+
- Correct Answer ----[string]-->|C |
- Student's Answer --[string]-->| O |
- | M |
- | P|
- +----+
- ---- |
- +---- \ <--------------------------[int] Current Score
- | / |
- | ---- V
- V +------+
- History of Scores --[array of int]-->| ANN |--[int]--> Confidence
- +------+
-
- The ``Current Score'' represents the correlation between the student's
- answer and the correct answer, and the ``Confidence'' value represents the
- state of the learning curve. I would like to use this confidence value to
- modulate the test dynamically, feeding back in such a way that confidence
- increases at the fastest rate possible.
-
- That is, I want to use the ANN to control a learning curve, the student's
- learning curve.
-
- Has anybody seen such an application before?
-
- Tia,
- Dave
- -----
- David Wuertele, Yasuda Lab, Electronic Engineering, Institute of Industrial Science,
- University of Tokyo. dave@windsor.iis.u-tokyo.ac.jp
-
-