A faceted approach to conceptualizing tasks in information seeking Yuelin Li a, * , Nicholas J. Belkin b,1 a School of Library and Information Science, The University of Southern Mississippi, 118 College Drive #5146, Hattiesburg, MS 39401, United States b School of Communication, Information and Library Studies, Rutgers University, 4 Huntington Street, New Brunswick, NJ 08901, United States article info Article history: Received 8 September 2007 Received in revised form 22 July 2008 Accepted 28 July 2008 Available online 14 September 2008 Keywords: Work task Search task Task classification Information-seeking behavior Information search behavior abstract The nature of the task that leads a person to engage in information interaction, as well as of information seeking and searching tasks, have been shown to influence individuals’ infor- mation behavior. Classifying tasks in a domain has been viewed as a departure point of studies on the relationship between tasks and human information behavior. However, pre- vious task classification schemes either classify tasks with respect to the requirements of specific studies or merely classify a certain category of task. Such approaches do not lead to a holistic picture of task since a task involves different aspects. Therefore, the present study aims to develop a faceted classification of task, which can incorporate work tasks and information search tasks into the same classification scheme and characterize tasks in such a way as to help people make predictions of information behavior. For this purpose, previous task classification schemes and their underlying facets are reviewed and dis- cussed. Analysis identifies essential facets and categorizes them into Generic facets of task and Common attributes of task. Generic facets of task include Source of task, Task doer, Time, Action, Product, and Goal. Common attributes of task includes Task characteristics and User’s perception of task. Corresponding sub-facets and values are identified as well. In this fashion, a faceted classification of task is established which could be used to describe users’ work tasks and information search tasks. This faceted classification provides a framework to further explore the relationships among work tasks, search tasks, and inter- active information retrieval and advance adaptive IR systems design. Published by Elsevier Ltd. 1. Introduction The effectiveness of information retrieval (IR) systems is measured based on how well users’ information problems are resolved and to what extent the information retrieved helps users to achieve their goals. Users’ information problems, and their contexts, are various and thus need different types of information support. However, a big disadvantage of current IR systems, including search engines and digital libraries, is that they typically offer support for only one type of information problem, and present search results based on the queries input regardless of users’ divergent tasks with different goals. As a result, users’ interactions in IR systems may be sub-optimal, and returned search results may not help users to achieve their goals. Therefore, it is necessary to explore how IR systems could adapt to a variety of users with different goals, contexts and types of information problems; that is, to contextualize and personalize interaction with IR systems. One aspect of characterizing information-seekers’ goals, contexts and information problems is to consider the tasks which have led them to engage in information-seeking behavior, and the tasks that they need to accomplish in information seeking and in information searching, respectively. Various researchers have demonstrated that information behaviors and desired 0306-4573/$ - see front matter Published by Elsevier Ltd. doi:10.1016/j.ipm.2008.07.005 * Corresponding author. Tel.: +1 601 266 4035; fax: +1 601 266 5774. E-mail addresses: Yuelin.Li@usm.edu (Y. Li), nick@belkin.rutgers.edu (N.J. Belkin). 1 Tel.: +1 732 932 7500x8271; fax: +1 732 932 6916. Information Processing and Management 44 (2008) 1822–1837 Contents lists available at ScienceDirect Information Processing and Management journal homepage: www.elsevier.com/locate/infoproman