Clustering of Brazilian legal judgments about failures in air transport service: an evaluation of different approaches

The paper presents different clustering approaches in legal judgments from the Special Civil Court located at the Federal University of Santa Catarina (JEC/UFSC). The subject is Consumer Law, specifically cases in which consumers claim moral and material compensation from airlines for service failures. To identify patterns from the dataset, we apply four types of clustering algorithms: Hierarchical and Lingo (soft clustering), K-means and Affinity Propagation (hard clustering). We evaluate the results based on the following criteria: (1) entropy and purity; (2) algorithm's ability in providing labels; (3) legal expert’s evaluation; and (4) experimental complexity. The results demonstrate that the most advantageous approach is Hierarchical Clustering, since it has the best entropy and purity numbers, as well as the least difficulty for the expert to analyze the clusters, and the least experimental complexity. The main contribution of the paper is to show the advantages and disadvantages of each approach, especially to identify labels in unstructured and non-indexed legal texts.

     

A direct algorithm for constrained variational problems in several dimensions

We introduce a heuristic, practical procedure to take into account, in an easy-to-implement way, point-wise constraints in a variational problem in several dimensions. In addition to showing a convergence result under suitable assumptions, we emphasize the flexibility of the method by using it for an optimal design problem either in conductivity or elasticity, where an optimal mixture of two materials is to be found in a given design domain, under pointwise constraints. We illustrate the performance of the algorithm with some simple test examples.     

Crystal Structure of Escherichia coli Agmatinase: Catalytic Mechanism and Residues Relevant for Substrate Specificity

Agmatine is the product of the decarboxylation of L-arginine by the enzyme arginine decarboxylase. This amine has been attributed to neurotransmitter functions, anticonvulsant, anti-neurotoxic, and antidepressant in mammals and is a potential therapeutic agent for diseases such as Alzheimer’s, Parkinson’s, and cancer. Agmatinase enzyme hydrolyze agmatine into urea and putrescine, which belong to one of the pathways producing polyamines, essential for cell proliferation. Agmatinase from Escherichia coli (EcAGM) has been widely studied and kinetically characterized, described as highly specific for agmatine. In this study, we analyze the amino acids involved in the high specificity of EcAGM, performing a series of mutations in two loops critical to the active-site entrance. Two structures in different space groups were solved by X-ray crystallography, one at low resolution (3.2 Å), including a guanidine group; and other at high resolution (1.8 Å) which presents urea and agmatine in the active site. These structures made it possible to understand the interface interactions between subunits that allow the hexameric state and postulate a catalytic mechanism according to the Mn²⁺ and urea/guanidine binding site. Molecular dynamics simulations evaluated the conformational dynamics of EcAGM and residues participating in non-binding interactions. Simulations showed the high dynamics of loops of the active site entrance and evidenced the relevance of Trp68, located in the adjacent subunit, to stabilize the amino group of agmatine by cation-pi interaction. These results allow to have a structural view of the best-kinetic characterized agmatinase in literature up to now.     

Efficient data partitioning for the GPU computation of moment functions

In our previous work, we have provided tools for an efficient characterization of biomedical images using Legendre and Zernike moments, showing their relevance as biomarkers for classifying image tiles coming from bone tissue regeneration studies (Ujaldón, 2009) [24]. As part of our research quest for efficiency, we developed methods for accelerating those computations on GPUs (Martín-Requena and Ujaldón, 2011)  and . This new stage of our work focuses on the efficient data partitioning to optimize the execution on many-cores and clusters of GPUs to attain gains up to three orders of magnitude when compared to the execution on multi-core CPUs of similar age and cost using 1 Mpixel images. We deploy a successive and successful chain of optimizations which exploit symmetries in trigonometric functions and access patterns to image pixels which are effectively combined with massive data parallelism on GPUs to enable (1) real-time processing for our set of input biomedical images, and (2) the use of high-resolution images in clinical practice.