Publications
Point Set Registration (PSR) algorithms have very different underlying theoretical models to define a process that calculates the alignment solution between two point clouds. The selection of a particular PSR algorithm can be based on the efficiency (time to compute the alignment) and accuracy (a measure of error using the estimated alignment). In our specific context, previous work used a CPD algorithm to detect and quantify change in spatiotemporal datasets composed of moving and shape-changing objects represented by a sequence of time stamped 2D polygon boundaries. Though the results were promising, we question if the selection of a particular PSR algorithm influences the results of detection and quantification of change. In this work we review and compare several PSR algorithms, characterize test datasets and used metrics, and perform tests for the selected datasets. The results show pyCPD and cyCPD implementations of CPD to be good alternatives and that BCPD can have potential to be yet another alternative. The results also show that detection and quantification accuracy change for some of the tested PSR implementations.
Index structures were often used to optimise fetch operations to external storage devices (secondary memory). Nowadays, this also holds for increasingly large amounts of data residing in main-memory (primary memory). Within this scope, this work focuses on index structures that efficiently insert, query and delete valid-time data from very large datasets. This work performs a comparative study on the performance of the Interval B+ tree (IB+ tree) and the Improved Interval B+ tree (I2B+ tree): a variant that improves the time-efficiency of the deletion operation by reducing the number of traversed nodes to access siblings. We performed an extensive analysis of the performance of two operations: insertions and deletions, on both index structures, using multiple datasets with growing volumes of data, distinct temporal distributions and tree parameters (time-split alpha and node order). Results confirm that the I2B+ tree globally outperforms the IB+ tree, since, on average, deletion operations are 7% faster, despite insertions requiring 2% more time. Furthermore, results also allowed to determine the key factors that augment the performance difference on deletions between both trees.
Information visualization commonly aids the understanding of the evolution of spatiotemporal phenomena. The current work proposes a novel approach to visually represent spatiotemporal phenomena based on the automated generation of static and interactive visual narratives that summarize the evolution of a spatiotemporal phenomenon. The visual narrative is composed of an interactive storyboard that consists of a set of frames that represent events of interest in the phenomenon. Towards corroborating the hypothesis that this approach would effectively and efficiently transmit the evolution of spatiotemporal phenomena, we conceptualized a visualization framework, identifying visual metaphors that map spatiotemporal transformations into visual content and defining the parameterization approaches for spatiotemporal features. We developed a functional prototype implementing the conceptual solution and presented issues encountered regarding visual clutter and parameterization. We conducted a user study based on a questionnaire which concluded that the proposed approach can be effective and efficient for understanding the evolution of these phenomena in terms of transformations for a subset of possible scenarios.
Index structures are fast-access methods. In the past, they were often used to minimise fetch operations to external storage devices (secondary memory). Nowadays, this also holds for increasingly large amounts of data residing in main-memory (primary memory). Examples of software that deals with this fact are in-memory databases and mobile device applications. Within this scope, this paper focuses on index structures to store, access and delete interval-based time-dependent (temporal) data from very large datasets, in the most efficient way. Index structures for this domain have specific characteristics, given the nature of time and the requirement to index time intervals. This work presents an open-source time-efficiency focused variant of the original Interval B+ tree. We designate this variant Improved Interval B+ tree (I2B+ tree). Our contribution adds to the performance of the delete operation by reducing the amount of traversed nodes to access siblings. We performed an extensive analysis of insert, range queries and deletion operations, using multiple datasets with growing volumes of data, distinct temporal distributions and tree parameters (time-split and node order). Results of the experiments validate the logarithmic performance of these operations and propose the best-observed tree parameter ranges.
Advanced Traveller Information Systems (ATIS) have seen a steady increase in popularity in the last decades among urban users. By providing travellers with relevant information, these systems have the ability to considerably improve traffic flow, but are limited by their penetration rate among the city's population. In this work we put forward HERMES 1 , a tool to facilitate the evaluation of road networks through simulation with different ATIS and with different levels of information percolation among users. The usefulness of such a tool stems from the quantification of the extent to which an ATIS can improve total travel time and road utilisation by means of experimenting with simulation scenarios. This tool operates at a mesoscopic level, in which single users may be represented independently from each other, nonetheless are rather considered to operate at their macroscopic interactions with the network. We also illustrate example use cases for the tool on the motorway network between the Portuguese cities of Lisbon and Porto.