Calendar - University of Houston
Skip to main content

[Defense] Density-based Frameworks for Spatio-Temporal Change Analysis

Tuesday, November 30, 2021

9:00 am - 10:00 am

In Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy
Karima Elgarroussi
will defend her proposal
Density-based Frameworks for Spatio-Temporal Change Analysis


Abstract

Due to the advances in technology, the number and volume of spatio-temporal datasets has increased. Traditional statistical methods for dealing with such data are becoming overwhelmed. The field of spatio-temporal data mining (STDM) emerged out of a need to create effective and efficient techniques in order to reveal interesting spatio-temporal patterns from such data. On the other hand, density functions have served as a valuable tool in data mining. However, the development of spatio-temporal data mining techniques based on density estimation functions still in its infancy. To address the need for more robust and versatile analysis tools, density-based frameworks for spatio-temporal mining are designed and implemented in this research using different datasets. In our preliminary work, we introduced a density-based spatio-temporal data analysis, change analysis and storytelling frameworks for emotion mapping and emotion change analysis. Our proposed approach segments first the input tweet dataset into batches based on a fixed size time window. Next, by generalizing existing kernel density estimation techniques, each batch is transformed into a weighted continuous function that takes positive and negative values. Then, a contouring algorithm is employed to obtain spatial polygons. A set of unary and binary change predicates are evaluated with respect to the set of spatial clusters in batches. Next, an emotion change graph whose nodes are the spatial clusters and whose edges capture the temporal relationships of different types between spatial clusters is constructed and mined based on story types. The frameworks were successfully applied to tweets collected in the state of New York in June 2014; the experimental results show that the framework can effectively discover interesting spatio-temporal patterns and analyze their change over time. Nowadays, there is a huge interest in mining collocation patterns which identifies groups of particular features that appear frequently close to each other in a geo-spatial map. As density functions have been demonstrated to be an efficient tool to model the distance-decay effect, we developed a novel density-based collocation mining framework which introduces collocation measures estimating the strength of a given collocation pattern in a particular location. They are defined by combining non-parametric density functions associated with individual features in the longitude-latitude space. This approach is generalized to mine spatio-temporal collocation patterns proposing two approaches: the batch-based approach and the 3d density-based approach. As we are interested in this research at georeferenced data, we cannot ignore the importance of GIS in supporting spatio-temporal modeling. For that, we are currently designing a new graph based spatio-temporal change analysis framework using density-based COPULA for mining purposes.


Tuesday, November 30, 2021
9:00 AM - 10:00 AM CT
Online via MSFT Teams

Dr. Christoph Eick, dissertation advisor

Faculty, students and the general public are invited.