The effect of variations of prior on knowledge tracing

Matti Nelimarkka; Madeeha Ghori

The effect of variations of prior on knowledge tracing

^*Tämän työn vastaava kirjoittaja

University of California at Berkeley

Tutkimustuotos: Artikkeli kirjassa/konferenssijulkaisussa › Conference article in proceedings › Scientific › vertaisarvioitu

3 Sitaatiot (Scopus)

62 Lataukset (Pure)

Abstrakti

Knowledge tracing is a method which enables approximation of a student's knowledge state using a Bayesian network for approximation. As the applications of this method increase, it is vital to understand the limits of this approximation. We are interested how well knowledge tracing performs when students' prior knowledge on the topic is extremely high or low. Our results indicate that the estimates become more erroneous when prior knowledge is extremely high (prior =0.90).

Alkuperäiskieli	Englanti
Otsikko	EDM 2014 Extended Proceedings
Alaotsikko	Proceedings of the Workshops held at Educational Data Mining 2014, co-located with 7th International Conference on Educational Data Mining (EDM 2014), London, United Kingdom, July 4-7, 2014
Toimittajat	Sergio Gutierrez-Santos, Olga C. Santos
Sivut	146-150
Sivumäärä	5
Tila	Julkaistu - 2014
OKM-julkaisutyyppi	A4 Artikkeli konferenssijulkaisussa
Tapahtuma	International Conference on Educational Data Mining - London, Iso-Britannia Kesto: 4 heinäk. 2014 → 7 heinäk. 2014 Konferenssinumero: 7

Julkaisusarja

Nimi	CEUR workshop proceedings
Vuosikerta	1183
ISSN (painettu)	1613-0073

Conference

Conference	International Conference on Educational Data Mining
Lyhennettä	EDM
Maa/Alue	Iso-Britannia
Kaupunki	London
Ajanjakso	04/07/2014 → 07/07/2014

Pääsy asiakirjaan

bkt20y_paper06Final published version, 282 KB

Muut tiedostot ja linkit

Link to publication in Scopus

Siteeraa tätä

Nelimarkka, M., & Ghori, M. (2014). The effect of variations of prior on knowledge tracing. teoksessa S. Gutierrez-Santos, & O. C. Santos (Toimittajat), EDM 2014 Extended Proceedings: Proceedings of the Workshops held at Educational Data Mining 2014, co-located with 7th International Conference on Educational Data Mining (EDM 2014), London, United Kingdom, July 4-7, 2014 (Sivut 146-150). (CEUR workshop proceedings; Vuosikerta 1183).

Nelimarkka, Matti ; Ghori, Madeeha. / The effect of variations of prior on knowledge tracing. EDM 2014 Extended Proceedings: Proceedings of the Workshops held at Educational Data Mining 2014, co-located with 7th International Conference on Educational Data Mining (EDM 2014), London, United Kingdom, July 4-7, 2014. Toimittaja / Sergio Gutierrez-Santos ; Olga C. Santos. 2014. Sivut 146-150 (CEUR workshop proceedings).

@inproceedings{6e147a06a9674bdaa9e5009edab24d6b,

title = "The effect of variations of prior on knowledge tracing",

abstract = "Knowledge tracing is a method which enables approximation of a student's knowledge state using a Bayesian network for approximation. As the applications of this method increase, it is vital to understand the limits of this approximation. We are interested how well knowledge tracing performs when students' prior knowledge on the topic is extremely high or low. Our results indicate that the estimates become more erroneous when prior knowledge is extremely high (prior =0.90).",

keywords = "Bayesian knowledge tracing, Parameter estimation, Personalization, Prior",

author = "Matti Nelimarkka and Madeeha Ghori",

year = "2014",

language = "English",

series = "CEUR workshop proceedings",

pages = "146--150",

editor = "Sergio Gutierrez-Santos and Santos, {Olga C.}",

booktitle = "EDM 2014 Extended Proceedings",

note = "International Conference on Educational Data Mining , EDM ; Conference date: 04-07-2014 Through 07-07-2014",

}

Nelimarkka, M & Ghori, M 2014, The effect of variations of prior on knowledge tracing. julkaisussa S Gutierrez-Santos & OC Santos (toim), EDM 2014 Extended Proceedings: Proceedings of the Workshops held at Educational Data Mining 2014, co-located with 7th International Conference on Educational Data Mining (EDM 2014), London, United Kingdom, July 4-7, 2014. CEUR workshop proceedings, Vuosikerta 1183, Sivut 146-150, International Conference on Educational Data Mining , London, Iso-Britannia, 04/07/2014.

The effect of variations of prior on knowledge tracing. / Nelimarkka, Matti; Ghori, Madeeha.
EDM 2014 Extended Proceedings: Proceedings of the Workshops held at Educational Data Mining 2014, co-located with 7th International Conference on Educational Data Mining (EDM 2014), London, United Kingdom, July 4-7, 2014. toim. / Sergio Gutierrez-Santos; Olga C. Santos. 2014. s. 146-150 (CEUR workshop proceedings; Vuosikerta 1183).

Tutkimustuotos: Artikkeli kirjassa/konferenssijulkaisussa › Conference article in proceedings › Scientific › vertaisarvioitu

TY - GEN

T1 - The effect of variations of prior on knowledge tracing

AU - Nelimarkka, Matti

AU - Ghori, Madeeha

N1 - Conference code: 7

PY - 2014

Y1 - 2014

N2 - Knowledge tracing is a method which enables approximation of a student's knowledge state using a Bayesian network for approximation. As the applications of this method increase, it is vital to understand the limits of this approximation. We are interested how well knowledge tracing performs when students' prior knowledge on the topic is extremely high or low. Our results indicate that the estimates become more erroneous when prior knowledge is extremely high (prior =0.90).

AB - Knowledge tracing is a method which enables approximation of a student's knowledge state using a Bayesian network for approximation. As the applications of this method increase, it is vital to understand the limits of this approximation. We are interested how well knowledge tracing performs when students' prior knowledge on the topic is extremely high or low. Our results indicate that the estimates become more erroneous when prior knowledge is extremely high (prior =0.90).

KW - Bayesian knowledge tracing

KW - Parameter estimation

KW - Personalization

KW - Prior

UR - http://www.scopus.com/inward/record.url?scp=84922295584&partnerID=8YFLogxK

M3 - Conference article in proceedings

AN - SCOPUS:84922295584

T3 - CEUR workshop proceedings

SP - 146

EP - 150

BT - EDM 2014 Extended Proceedings

A2 - Gutierrez-Santos, Sergio

A2 - Santos, Olga C.

T2 - International Conference on Educational Data Mining

Y2 - 4 July 2014 through 7 July 2014

ER -

Nelimarkka M, Ghori M. The effect of variations of prior on knowledge tracing. julkaisussa Gutierrez-Santos S, Santos OC, toimittajat, EDM 2014 Extended Proceedings: Proceedings of the Workshops held at Educational Data Mining 2014, co-located with 7th International Conference on Educational Data Mining (EDM 2014), London, United Kingdom, July 4-7, 2014. 2014. s. 146-150. (CEUR workshop proceedings).

The effect of variations of prior on knowledge tracing

Abstrakti

Julkaisusarja

Conference

Pääsy asiakirjaan

Muut tiedostot ja linkit

Sormenjälki

Siteeraa tätä