Learning to predict closed questions on Stack Overflow
Učënye zapiski Kazanskogo universiteta. Seriâ Fiziko-matematičeskie nauki, Uchenye Zapiski Kazanskogo Universiteta. Seriya Fiziko-Matematicheskie Nauki, Tome 155 (2013) no. 4, pp. 118-133
Voir la notice du chapitre de livre provenant de la source Math-Net.Ru
The paper deals with the problem of predicting whether the user's question will be closed by the moderator on Stack Overflow, a popular question answering service devoted to software programming. The task along with data and evaluation metrics was offered as an open machine learning competition on Kaggle platform. To solve this problem, we employed a wide range of classification features related to users, their interactions, and post content. Classification was carried out using several machine learning methods. According to the results of the experiment, the most important features are characteristics of the user and topical features of the question. The best results were obtained using Vowpal Wabbit – an implementation of online learning based on stochastic gradient descent. Our results are among the best ones in overall ranking, although they were obtained after the official competition was over.
Keywords:
community question answering systems
Mots-clés : large-scale classification, question classification.
Mots-clés : large-scale classification, question classification.
@article{UZKU_2013_155_4_a11,
author = {G. Lezina and A. Kuznetsov and P. Braslavski},
title = {Learning to predict closed questions on {Stack} {Overflow}},
journal = {U\v{c}\"enye zapiski Kazanskogo universiteta. Seri\^a Fiziko-matemati\v{c}eskie nauki},
pages = {118--133},
publisher = {mathdoc},
volume = {155},
number = {4},
year = {2013},
language = {en},
url = {http://geodesic.mathdoc.fr/item/UZKU_2013_155_4_a11/}
}
TY - JOUR AU - G. Lezina AU - A. Kuznetsov AU - P. Braslavski TI - Learning to predict closed questions on Stack Overflow JO - Učënye zapiski Kazanskogo universiteta. Seriâ Fiziko-matematičeskie nauki PY - 2013 SP - 118 EP - 133 VL - 155 IS - 4 PB - mathdoc UR - http://geodesic.mathdoc.fr/item/UZKU_2013_155_4_a11/ LA - en ID - UZKU_2013_155_4_a11 ER -
%0 Journal Article %A G. Lezina %A A. Kuznetsov %A P. Braslavski %T Learning to predict closed questions on Stack Overflow %J Učënye zapiski Kazanskogo universiteta. Seriâ Fiziko-matematičeskie nauki %D 2013 %P 118-133 %V 155 %N 4 %I mathdoc %U http://geodesic.mathdoc.fr/item/UZKU_2013_155_4_a11/ %G en %F UZKU_2013_155_4_a11
G. Lezina; A. Kuznetsov; P. Braslavski. Learning to predict closed questions on Stack Overflow. Učënye zapiski Kazanskogo universiteta. Seriâ Fiziko-matematičeskie nauki, Uchenye Zapiski Kazanskogo Universiteta. Seriya Fiziko-Matematicheskie Nauki, Tome 155 (2013) no. 4, pp. 118-133. http://geodesic.mathdoc.fr/item/UZKU_2013_155_4_a11/