Exploring long-term behavior patterns in a book recommendation system for reading

Introduction Book recommendation between students has the potential to enhance reading interest and ability (Chien, Chen, Ko, Ku, & Chan, 2015). The recommendation process encourages students to share and discuss what they have read, thus helping to create and foster a good reading atmosphere within the class (Larson, 2009). A book recommendation system can provide students with an opportunity to review what they have read by retelling, based on their own individual understanding and perspectives, which can greatly improve their literacy skills, reading comprehension and recollection (Carico, Logan, & Labbo, 2004). However, the benefits of a book recommendation system are reliant upon extensive and frequent student interaction, which requires overcoming the difficulties of organizational complexity and limited interaction. The former relates to the problem of class time allocation. It is difficult for teachers in the classroom to clear space on the agenda for a book recommendation activity and ensure that all students have an equal chance to express their opinions. The latter refers to the time needed in order to reinforce the effect of social interaction. In small class size elementary classrooms, there might not be enough student interaction for peer-to-peer book recommendation. Strategies proposed to overcome these two difficulties include activities designed to reduce the complexity and enhance the efficiency of social interaction, such as book talks, storytelling, and dramas (Pilgreen, 2000; Gardiner, 2005; Atwell, 2007). Although these activities can greatly help to facilitate classroom book recommendation, they still require the teacher to expend a lot of effort on logistical organization, such as setting up tables, group allocation, and maintaining classroom order. Recent developments in information technology however, have opened the door for teachers to provide tech-support support for the development of communication tools and book recommendation systems (Hamilton & Cherniavsky, 2006, Larson, 2008) in the classroom, include message board discussions (Wolsey, Biesenbach-Lucas, & Meloni, 2004), blogs (Huffaker, 2005; Ray, 2006), and social networking systems. With the support of technology, teachers can efficiently plan activities to help students express their opinions to their peers, both in and out of school time (Hancock, 2008). These technology-supported activities can serve as a stage, on which students can express ideas, perspectives, and thoughts to their classmates, and help to foster the development of a learning community for reading and book recommendation (Wolsey, Biesenbach-Lucas, & Meloni, 2004; Larson, 2009). One example is to use potential of technology for promoting the behavior of book recommendation. The MyBookstore system (Chien, Chen, Ko, Ku, & Chan, 2011), which incorporates incentive models into a recommendation system, is designed to help students describe their favorite books and recommend them to their classmates. Although a previous study has demonstrated that such a system can enhance student learning in terms of word usage and reading perception (Chien et al., 2015), little attention was paid to investigating the students' behavior patterns over a long period of time. Investigating student behavior when using such a book recommendation system is critical because not only do the findings enrich our understanding of how an effective system works, but also sheds light on how to stimulate and maintain positive behaviors. The purpose of this study is thus to conduct a long-term empirical study using the My-Bookstore system as an example. In addition, to acquire a more comprehensive understanding of student behavior, we not only analyze the frequency of certain behaviors, but also trace their sequence. In this way, subtle changes in behavior can be revealed, even if the frequencies remain similar. The two research questions to be answered in this study are: (1) What are the frequent behaviors and behavioral patterns of students participating in a book recommendation system? …