Tadava : visual analytics architecture for large table-based datasets

Visual Analytics provide the user with tools to process data in a very intuitive way. One of the challenges Visual Analytics faces nowadays is the need to represent big amounts of information in a way that the user can explore. This large amount of data can not be managed by conventional machines and must be partitioned or underrepresented. Usually, visual analytics applications or widgets work with one single machine due to the requirements of low latency and quick interactions. This article presents Tadava, a visual analytics architecture based on representative sampling for large table-based datasets. Datasets samples are generated using systematics sampling and random sampling in order to obtain a collection of data from the original dataset. Experiments are made to determine the best sampling method to be used in datasets of different sizes, varying the step expected between samples. Tadava is built as a backend architecture for Navio, an interactive visualization widget for summarizing, exploring and navigating large datasets, and provides the widget with the capacity to manage table-based datasets between 400MB and 4GB.

• Autores: Ortiz Román, Juan Camilo
• Fecha de publicación: 2018
• Institución: Universidad de los Andes
• Repositorio: Séneca: repositorio Uniandes
• Idioma: eng
• OAI Identifier: oai:repositorio.uniandes.edu.co:1992/34912
• Acceso en línea: http://hdl.handle.net/1992/34912
• Palabra clave: Analítica visual - Investigaciones - Estudio de casos; Visualización de la información - Investigaciones; Arquitectura de software - Investigaciones; Software de aplicación - Investigaciones - Estudio de casos; Ingeniería
• Rights: openAccess
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/