@misc{374, keywords = {Visual analytics, Matrix Visualization}, author = {Bilal Alsallakh}, title = {Visual Analytics of Large Multivariate Matrix Data}, abstract = {Matrices are generic data structures that are used to record and summarize data in a variety of scientific and business domains. Large data matrices are potentially rich of information and patterns that are interesting to detect and analyze. Visual analysis of matrix data is typically performed using common visualization methods such as heat maps and parallel coordinates. However, these methods impose limitations on the scalability and on the insights that can be gained from the data. This research proposal aims to address a class of matrix data that have few (tens) columns and a large number (thousands) of rows. After giving examples from which such data arise and discussing related work, a Visual Analytics approach is proposed for analyzing this class of matrix data. The benefits of this approach are demonstrated by means of three use cases that share similar data structures and analysis goals, but have different nature and semantics of the data. Lastly, the proposal briefly describes the results reached so far with one particular type of data. It then outlines the research planned to generalize these results to all matrix data of the class mentioned above. }, year = {2012}, journal = {Poster: IEEE VisWeek Doctoral Colloquium}, month = {10/2012}, address = {Seattle}, url = {http://publik.tuwien.ac.at/files/PubDat_212124.pdf}, note = {Poster: IEEE VIS Doctoral Colloquium }, }