Ziyuan Lin · Fri, 24.03.2017
The 12th International Conference on Information Visualization Theory and Applications (IVAPP 2017) was hosted at Porto, Portugal from 27/02/2017 to 01/03/2017. It is part of the 12th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2017). Topics of information visualization that IVAPP focuses on range from visualization system for specific data to general and abstract information visualization theories or models, including methodologies coming from dimensionality reduction, a familiar topic for the machine learning community.
My paper Parallel Coordinate Plots for Neighbor Retrieval, coauthored with my instructor Jaakko Peltonen, is accepted to IVAPP 2017. It proposes a kind of parallel coordinate plots (PCPs) optimized for a recently studied user task called neighbor retrieval. In the optimized PCPs, axes with similar neighbor relations are placed nearby and vice versa, minimizing the retrieval error in neighbor retrieval. The paper falls into the submission category of “abstract information visualization”. The presentation of the paper was given at 10:45—11:15, 28/02/2017.
The paper will be included in my underwriting thesis Exploratory Data Analysis via Neighbor Embedding and Retrieval. The thesis mainly discusses about rigorous visualizations according to information retrieval measures obtained from neighbor embedding – a family of dimensionality algorithms. As one of the main themes, the thesis considers optimized visualizations for user tasks of neighbor retrieval beyond simple scatter plots. This paper will be part of chapter 5, among other visualization techniques that support neighbor retrieval. The draft of the thesis is expected to come out in June 2017.