Context-aware Driven Spatio-Temporal Data Adaptive Visualizations In Mobile Environment

The progress in the fields of mobile Internet and’3S’technologies, together with the portable use of handled devices leads to new opportunities and challenges for mobile spatial information services (MSIS). And it is becoming an important part of our dairy life and work.The visual information on mobile devices is the most direct communicative manner with the end user. Moreover, the simple and intuitive visualizations can help the user acquire the necessary information more efficiently. However, developing spatio-temporal data adaptive visualization methods for MSIS is an essential and complex issue, particularly on the small display devices. The information visualization on these devices still have many deficiencies because of the limitations of software/hardware configurations and traditional visualization theory, the influences of the new environment, and the user’s new demands.Taking vehicle navigation services or mobile guided systems as an example, a lot of the disadvantages are as follows:(ⅰ) they are subjected to limitations like limited memory capacities and small displays including low resolution so that the traditional data organization and visualization theories cannot cope with the transmission, expression and interpretation of the information,(ⅱ) users appear to have difficulties in focusing their attention on the display for a long time in order to interactive operation or acquire the prompt information in mobile environment. Meanwhile, there is an increasing trend for MSIS to provide user with personalized services. And what’s more, the user pays more attention to the information relevant to them, whereas the traditional navigation systems are lack of filtering the useless information or recommending the useful one,(ⅲ) although recently the visualization based on context-aware has become a trend along with the development of sensing technology, few vehicle navigation systems or mobile guided systems can provide this form of services.As mentioned above, the information visualization on mobile device still has to face many difficulties, such as the inappropriate scale, lots of redundant information, response latency of the situation, and hard to recognized, etc. Motivated by the limitations of existing approaches, we propose a framework of context-aware driven spatio-temporal data adaptive visualizations (STDAV) for mobile environment in this thesis. And this research contributes to the domain of STDAV in the following ways:1. The research background and significance of this thesis are introduced and discussed. First of all, an in-depth analysis of the challenges and opportunities is given in terms of communical facilities, display devices, market and policies. Then the significance of developing the spatio-temporal data adaptive visualizations for MSIS is dicussed. Based on a thorough review of the literature, this thesis describes the state-of-the-art of adaptive geovisualization and identifies major issues in the previous research from aspects of technological framework, information domain, visualization domain, user interface domain, and technology domain. After that this thesis defines the objectives and research questions. 2. In general, this thesis proposes a dynamic requirement oriented framework for adaptive spatio-temporal data visualizations. It elaborates the research trend of spatio-temporal data visualizations in a changing technological context, and highlights the user-centric design for geovisualization. Focusing on applications in Location Based Services (LBS), this research firstly makes a survey of user requirements (both vehicle navigation and smart phone) on geo-information and representations. By investigating the interaction between users, devices and information from the point of view in information communication, we identify the factors that may influence the use and communication of spatial information in a mobile context. We also discuss how these factors correspond to the user requirements. Finally, a workflow is designed with design principle discussed, in order to better formulate our adaptive visualization model.3. The innovative notion of’geographic relevance’that relates user information needs and information retrieval and representation techniques is adopted in this research. This research firstly introduces the origin and definition of the term and the rules under which it is used. Based on classifications of relevant rules in related fields, this thesis proposes the scenario for geographic relevance, and identifies the relevance between information and scenario, and the relevance between information units. Then, this thesis discuss the global calculative method of the relevance between the user and information, or information amd information, by means of Stevens’scale of measurement theory, including the nominal scale, ordinal scale, interval scale and ratio scale. Finally, this thesis discusses specific measures on various kinds of relevance.4. Based on the concept of geographic relevance, this research proposes an approach to visualizing information reconstruction. It studies in more detail the key techniques for reconstruction, which includes (ⅰ) integration of user context into geographic context,(ⅱ) construct the spatio-temporal area of intrest (STAOI) considering the importance of inforamtion and time of intrest (TOI) in application, and (ⅲ) relevance degree of spatio-temporal data distinguishing between simple, complex and dynamic relevance. Then, we thereafter discuss the procedure for reconstructing information for spatio-temporal data adaptive visualizations, where it is implemented by measuring contextual similarity, selecting the adaptive patterns, calculating user’s dynamic degree of focus (DDOF) considering the STAOI, and optimizing the reconstructing information using local relevance.5. Later several adaptive strategies are designed to visualize spatio-temporal data take into the context accout. We demonstrate the adaptive rules for presentation in terms of cognitive efficiency, geographic relevance, aesthetic quality and usability, and dicuss their use in a mobile map. The main merits here are adaptive visual region identification and adaptive representation of point of interest (POI) and road networks. Further, we find factors that influence the identification of visual regions on mobile maps, based on which we propose an adaptive scaling technique on top of’the optimal user response distance’. Besides, we derive a specific map load model by theorical derivation and user’s experimental data, based on comparing the difference of map display on mobile devices, papers and computer screens. This can be used to guide the selection of road segments adaptively. Finally, this research also discusses the possibility of collaborative visualization of road network data and POI using Gestalt theory.6. Finally, key concepts of the proposed approach are implemented as a prototype system to demonstrate their potential and feasibility in real applications. This research uses road network data and POI of Wuhan, China to evaluate our approach to adaptive geovisualization. Meanwhile, criteria from the visual attention and saliency map are used for this evaluation in the before-and-after adaptive visualization.