The research of discovery management for effective decision making has long resided in the realm of intelligent systems, organizational research, and social science. Recently, it has been receiving more attention from the Applied analytics community. This chapter presents an in-depth survey of the related work in these areas. The survey begins with theories and empirical studies. They serve as the theory foundation and design guidelines of this work. Then, we present the state-of-the-art in Applied analytics and discuss the limitations.
Multidimensional Applied Analytics Exploration
Applied analytics is an emerging research area that targets the grand challenge of analyzing massive amounts of data. It combines techniques from multi-disciplinary fields, such as statistics information, machine learning, and cognitive psychology, for facilitating analytical reasoning. Among the motivations of the generation of this field, the need for analyzing large-scale multidimensional datasets is among the most significant ones since these datasets are standard in many application domains such defense, health, governance, business, and cyberspace. The discovery management for multidimensional data and explore an Applied exploration paradigm to facilitate multidimensional data exploration (Siemens, 2013). A number of techniques can be used in the proposed paradigm for exploring multidimensional data and generating discoveries. For example, automatic knowledge discovery techniques, such as subspace clustering algorithms, k-nearest neighbor search algorithms, and k-nearest match algorithms, can be used to partition a high dimensional data space into multiple smaller divisions. As meaningful divisions are constructed, they can be Appliedly explored via less scalable Appliedization.
Discovery management and decision making are widely studied in areas such as intelligent systems and organizational research. The proposed work has been inspired by various efforts from those areas. For example, the knowledge evolves dynamically depending on the context. Such dynamic nature requires information workers to effectively manage their knowledge, such as capturing the knowledge, categorizing and linking information corresponding to the knowledge, and presenting them in a meaningful way. However, computer simulations to examine the role and interrelationship between search processes that were forward-looking based on the actors' cognitive map of action-outcome linkages, and those that were backward-looking, or experience-based. In sensemaking research, the sensemaking is a mixture of retrospect and prospect. These efforts provide a solid theory foundation to our “look forward and look backward” paradigm for dynamic knowledge construction.
In social and organizational research, researchers have investigated how management activities benefit collaboration in a variety of collaborative tasks, such as emergency task management, tactical operations planning, and collaborative information synthesis (Ferguson & Shum, 2012). Often, collaborative workers come into collaboration only having completed their own individual work. They are unaware of what has been done and found by others. Therefore, collaborative workers need to manage and share their individual work to reach a common ground of the collaboration.
Learning Activities and Supporting Resources
Actancial Matrix: characters or actants.
Actants are autonomous, independent and capable of action units. Predicates represent the actions of actants, are subordinate to them and depend on them to exist. An example of actants and predicates: actants: Predicates: Fight Community Landowners ...