Construction of Hierarchical Cognitive Academic Map
Nowadays, mobile devices have been considered as a new platform for information services, and have been widely used in many fields. In mobile application services, the processing and representation of data is a key issue which has a great impact on the service quality. Knowledge map is regarded as an effective method and has been widely utilized in mobile devices. However, traditional knowledge maps employed in mobile devices are subject to a lack of cognition characteristics, which results in corresponding information services' being unable to match the users' cognition level, thus affecting the quality of services. In this paper, we propose a hierarchical cognitive academic map (HCAM) for the specific academic domain application background. HCAM can meet the needs of three basic levels of Bloom's cognition taxonomy model by distinguishing the academic attributes of nodes and relations between nodes. First, academic concepts are the basic units in HCAM and are classied into research object concepts and method/technique concepts, which meet the human's remembering cognition levels. Second, HCAM provides the implementation and collaboration relation between concepts, which satises the human's applying and understanding cognition levels. Third, technique/method concepts are organized in the form of hierarchical structure from the top down of which concepts' specificity for the domain get higher and higher. In addition, Bayesian rose tree clustering is adopted in the construction of this hierarchical structure and acquiring the cognition depth for each concept. Furthermore, experiments on information retrieval field and data mining field are performed to demonstrate the effectiveness and cognition characteristics of HCAM.
Knowledge representation, cognition, entropy, clustering algorithms, Bayesian Rose Tree.