The Web-revolution in publishing and reading is rapidly increasing the volume of online textbooks. Nowadays, for most of the subjects, a selection of online textbooks is available. Such an abundance leads to an interesting opportunity: if a student does not like how a primary textbook presents a particular topic s/he can always access its alternative (e.g. more detailed or advanced) presentation elsewhere. Modern e-learning environments could better support access to different versions of instructional material by generating intelligent links between the textbooks sections that present similar topics and concepts. This paper reports an attempt to investigate the problem of fine-grained intelligent linking of online textbooks based on the probabilistic topic modeling technology. Using collections of textbooks in two domains (Elementary Algebra and Information Retrieval), we have demonstrated that intelligent linking based on probabilistic topic models produces a much better modeling quality than traditional term-based approaches.
%0 Book Section
%1 citeulike:12891852
%A Guerra, Julio
%A Sosnovsky, Sergey
%A Brusilovsky, Peter
%B Scaling up Learning for Sustained Impact
%D 2013
%E Hernández-Leo, Davinia
%E Ley, Tobias
%E Klamma, Ralf
%E Harrer, Andreas
%I Springer Berlin Heidelberg
%K electronic-textbook
%P 125--138
%R 10.1007/978-3-642-40814-4_11
%T When One Textbook Is Not Enough: Linking Multiple Textbooks Using Probabilistic Topic Models
%U http://dx.doi.org/10.1007/978-3-642-40814-4_11
%V 8095
%X The Web-revolution in publishing and reading is rapidly increasing the volume of online textbooks. Nowadays, for most of the subjects, a selection of online textbooks is available. Such an abundance leads to an interesting opportunity: if a student does not like how a primary textbook presents a particular topic s/he can always access its alternative (e.g. more detailed or advanced) presentation elsewhere. Modern e-learning environments could better support access to different versions of instructional material by generating intelligent links between the textbooks sections that present similar topics and concepts. This paper reports an attempt to investigate the problem of fine-grained intelligent linking of online textbooks based on the probabilistic topic modeling technology. Using collections of textbooks in two domains (Elementary Algebra and Information Retrieval), we have demonstrated that intelligent linking based on probabilistic topic models produces a much better modeling quality than traditional term-based approaches.
@incollection{citeulike:12891852,
abstract = {{The Web-revolution in publishing and reading is rapidly increasing the volume of online textbooks. Nowadays, for most of the subjects, a selection of online textbooks is available. Such an abundance leads to an interesting opportunity: if a student does not like how a primary textbook presents a particular topic s/he can always access its alternative (e.g. more detailed or advanced) presentation elsewhere. Modern e-learning environments could better support access to different versions of instructional material by generating intelligent links between the textbooks sections that present similar topics and concepts. This paper reports an attempt to investigate the problem of fine-grained intelligent linking of online textbooks based on the probabilistic topic modeling technology. Using collections of textbooks in two domains (Elementary Algebra and Information Retrieval), we have demonstrated that intelligent linking based on probabilistic topic models produces a much better modeling quality than traditional term-based approaches.}},
added-at = {2018-03-19T12:24:51.000+0100},
author = {Guerra, Julio and Sosnovsky, Sergey and Brusilovsky, Peter},
biburl = {https://www.bibsonomy.org/bibtex/2d2b8b67dd90323c7de654142081a2cfc/aho},
booktitle = {Scaling up Learning for Sustained Impact},
citeulike-article-id = {12891852},
citeulike-linkout-0 = {http://dx.doi.org/10.1007/978-3-642-40814-4_11},
citeulike-linkout-1 = {http://link.springer.com/chapter/10.1007/978-3-642-40814-4_11},
doi = {10.1007/978-3-642-40814-4_11},
editor = {Hern\'{a}ndez-Leo, Davinia and Ley, Tobias and Klamma, Ralf and Harrer, Andreas},
interhash = {c4d115a20feaeceb7c06facc6da20169},
intrahash = {d2b8b67dd90323c7de654142081a2cfc},
keywords = {electronic-textbook},
pages = {125--138},
posted-at = {2014-01-05 21:02:59},
priority = {0},
publisher = {Springer Berlin Heidelberg},
series = {Lecture Notes in Computer Science},
timestamp = {2018-03-19T12:24:51.000+0100},
title = {{When One Textbook Is Not Enough: Linking Multiple Textbooks Using Probabilistic Topic Models}},
url = {http://dx.doi.org/10.1007/978-3-642-40814-4_11},
volume = 8095,
year = 2013
}