Learning to read efficiently and effectively is emphasized at the elementary and high school levels. Finding books that children/youth are interested in reading, however, is a non-trivial task due to the diversity of topics and different readability levels covered in the huge volume of books available these days. Ideally, K-12 students can turn to book recommenders which suggest books that match their interests. However, since the preferences and reading levels of these students vary from one grade to another, books suggested by existing recommenders, which ignore the literacy skills and the personal interests of their users, may be unsuitable for the targeted audience. In this paper, we present additional design issues that should be applied in developing a book recommender based on BReK12, our previously-proposed book recommender for K-12 users, to further enhance the quality of its recommendations. BReK12, which performs content and readability analysis to identify books potentially appealing to its users, is extended to incorporate (i) a multi-criteria analysis that studies its users' complex and diverse interests and (ii) an enhanced readability-detection tool that determines precisely the readability levels of books which match the literary skills of its users.
[1]
Chun Chen,et al.
Document recommendation in social tagging services
,
2010,
WWW '10.
[2]
Maria Soledad Pera,et al.
BReK12: a book recommender for K-12 users
,
2012,
SIGIR '12.
[3]
B. Ireland.
National Council for Curriculum and Assessment
,
2005
.
[4]
Gediminas Adomavicius,et al.
Multi-Criteria Recommender Systems
,
2011,
Recommender Systems Handbook.
[5]
Yiu-Kai Ng,et al.
ReadAid: A Robust and Fully-Automated Readability Assessment Tool
,
2011,
2011 IEEE 23rd International Conference on Tools with Artificial Intelligence.
[6]
Robert E. Slavin,et al.
Effective Reading Programs for the Elementary Grades: A Best-Evidence Synthesis
,
2009
.