Interest-guided Exploration of Large Information Space

1 online resource (126 pages) : PDF

2018

University of North Carolina at Charlotte

With the coming of the digital era, information has never been as massive and complicated as it is nowadays. New layout strategies, visual encoding methods, and exploration mantras have been developed in the visualization area to ``squeeze a billion records into a million pixels''. However, the result usually loses the intuitiveness and appealingness that make visualization effective in the first place. This dissertation presents an alternative approach to address the conflict between the volume of data and the scalability bottleneck of visualization. Instead of showing everything at once, it moves analysts close to their datasets to see the task-relevant part only. In this way, only a handful of information is rendered on screen at any time, and it becomes practicable to rely on visual presentations intuitive and easy to learn. The challenge is how to navigate users to find their desired information island in a massive sea, especially when the explorer may not have a crystal clear task. It is also expected that an exploration system should encourage users to discover unexpected yet useful information and build a thorough understanding of the datasets as a result. My mantra has three steps: land, modify and shift. First, a user lands in an information space via a user-given center. The system automatically extracts and visually presents information around it. Second, the user modifies the visualization to match his/her exploration intent better. Third, the user shifts the center as his/her exploration intent changes during the analysis. Flexible and intuitive visual interfaces and interactions are relied on to raise awareness and navigate exploration throughout the process.This dissertation presents the mantra and illustrates it through three distinct projects. For each project, the challenges, design concerns, implementation details, and evaluations are reported. They reveal the shared features, benefits, and limitations of the mantra. At the end of the dissertation, I will summarize a design guideline, features, benefits, and limitations of the mantra.

doctoral dissertations

Information technology
Information science

Ph.D.

Data Exploration
Recommendation
Visualization

Computer Science

Yang, Jing

Lu, Aidong
Fan, Wei
Subramanian, Kalpathi

Thesis (Ph.D.)--University of North Carolina at Charlotte, 2018.

This Item is protected by copyright and/or related rights. You are free to use this Item in any way that is permitted by the copyright and related rights legislation that applies to your use. For other uses you need to obtain permission from the rights-holder(s). For additional information, see http://rightsstatements.org/page/InC/1.0/.

Copyright is held by the author unless otherwise indicated.

Hu_uncc_0694D_11918

http://hdl.handle.net/20.500.13093/etd:1512