Interactive visualization for spatial data exploration
- Authors
- Noboa Morales, Tatiana Karina
- Format
- MasterThesis
- Status
- publishedVersion
- Description
The present work has the objective of constructing a software platform that combines appropriate libraries and technologies to interactively apply visual perception of geographical information, which will encourage the exploration, observation and validation of hypotheses with the objective of gaining an intuitive knowledge of a given subject when examining sets of data and techniques over this groups, besides, the product of the present work aim is to aid the comprehension of data exploration techniques as data diversification and query refinement techniques. To achieve this goal the stated platform combines various web technologies and services that allows the displaying of data exploration results in a spatial context. The functioning of the software environment is demonstrated in a web interactive platform that acting over two databases of geo-spatial information shows the behavior of max-min and max-sum dispersion methods for data diversification, and similarity aware query refinement SAQR techniques. Finally, the mentioned techniques are show-cased by solving specified real life scenarios, the first one illustrates the evenly distribution of cities among a rectangular area or country according to max-min diversification technique and the boundary distribution according to maxsum diversification technique, this could guide a tourist on visiting diverse cities on a determined country for example. The second one illustrates SAQR method of query refinement, which specifically aims to guide police chiefs in allocating zones of operation for their under command officers using a real, historical dataset of crime incidents of the city of San Diego, CA in USA.
- Publication Year
- 2014
- Language
- eng
- Topic
- BASES DE DATOS
DATA MINING
DATA DIVERSIFICATION
WEB VISUALIZATION
QUERY REFINEMENT
- Repository
- Repositorio SENESCYT
- Rights
- openAccess
- License