OpenScience
Team: Did you See this email request code for one of my paper. Your papers need to point to the data and code you did(!!!!!). Make sure your next paper has something that points to a stable web page that has your code and data. You can dump others stuff onto the page that has more data (say if your main paper is cut down to just a 3 page poster, use your web page to store that extra text. You can also post the longer version you submitted at that web page so its not extra work for you to make this web page.
I feel very strongly on this. So much so I refuse to have my name on any paper that does not do this. I strongly believe in making sure out lab does OPEN science. If we make errors I want others to be able to discover this. It's intellectually dishonest/sloppy to not do so, particularly given how you have heard me beat up on companies for not doing things like this I would be a hypocrite if I adopted any different policy. People should not have to email authors to get this information--think about the number of times someone decides not to email.
BUT OPEN Science is in your interests in a variety of ways. For instance, the reviewers will love you for it, so your paper is more likely to be accepted if your make your work open. They are willing to overlook your lack of detail in what is going on if they know you have given out your code. Also you will get more citations as well. There is a better chance you will get someone to build upon your work. And if some finds a bug in your work and write a paper about it, that is a citation as well!
Neil
---------- Forwarded message ----------
From: Benjamin Valdes <bvaldesa@itesm.mx>
Date: Mon, Nov 3, 2014 at 1:59 PM
Subject: information on “Learning Bayesian Knowledge Tracing Parameters with a Knowledge Heuristic"
To: nth@wpi.edu
Dear Dr. Heffernan,
I’m a Phd Student from Mexico working on reusability and portability of educational software components. I’m currently performing an analysis on user models and detectors used in the AIED, ITS and EDM communities to see which models or detectors could be reused and applied in other systems. I read your paper:
W. J. Hawkins, N. T. Heffernan, and R. S. J. D. Baker, “Learning Bayesian Knowledge Tracing Parameters with a Knowledge Heuristic,” in Proceedings of the 12th International Conference on Intelligent Tutoring Systems, 2014, pp. 150–155..
Could point me in the direction of where I can find the following details of the BKT model you proposed:
Availability: Is the model, or the system containing it, available for others? Is authorization for access required?
Portability: Can the model be used "as it is" in applications as it is?
Best regards,
--
Benjamín Valdés, MTI
Profesor de Cátedra
Departamento de Sistemas Computacionales,
Tecnológico de Monterrey
Campus Querétaro
SISTEMA TECNOLÓGICO DE MONTERREY
(+52 ) 44 2238-3278