Sense representations for Portuguese: experiments with sense embeddings and deep neural language models

Sense representations have gone beyond word representations like Word2Vec, GloVe and FastText and achieved innovative performance on a wide range of natural language processing tasks. Although very useful in many applications, the traditional approaches for generating word embeddings have a strict drawback: they produce a single vector representation for a given word ignoring the fact that ambiguous words can assume different meanings. In this paper, we explore unsupervised sense representations which, different from traditional word embeddings, are able to induce different senses of a word by analyzing its contextual semantics in a text. The unsupervised sense representations investigated in this paper are: sense embeddings and deep neural language models. We present the first experiments carried out for generating sense embeddings for Portuguese. Our experiments show that the sense embedding model (Sense2vec) outperformed traditional word embeddings in syntactic and semantic analogies task, proving that the language resource generated here can improve the performance of NLP tasks in Portuguese. We also evaluated the performance of pre-trained deep neural language models (ELMo and BERT) in two transfer learning approaches: feature based and fine-tuning, in the semantic textual similarity task. Our experiments indicate that the fine tuned Multilingual and Portuguese BERT language models were able to achieve better accuracy than the ELMo model and baselines.

     

An International Collaborative Study on Trypsin Inhibitor Assay for Legumes,Cereals, and Related Products

For determining trypsin inhibitor activity (TIA) in soy products, the American Oil Chemists' Society (AOCS) Method Ba 12-75 has been used. It measures differences in absorbance at 410 nm of bovine trypsin activity toward a synthetic substrate (Nα-benzoyl-DL-arginine-p-nitroanilide) in the absence and presence of an inhibitor. Recently, a significantly improved method was developed (JAOCS, 2019, 96:635–645), featuring 5 mL of total assay volume, enzyme-last sequence, and single inhibitor level in duplicate. It is proposed as the AOCS Method Ba 12a-2020. As a part of the AOCS method approval process, a collaborative study involving 12 international laboratories was conducted to evaluate the performance of the proposed method. The study involved measuring TIA in 10 selected test samples plus a blind duplicate. They included soybeans, pulses, cereals, and their processed products (flours, concentrates, and isolates). After rigorous statistical treatment of the data, only three outliers were removed from the data of two samples. Repeatability relative standard deviations (RSDr) for the 11 samples ranged from 0.99% to 5.52%. Reproducibility RSD (RSD_R) ranged from 7.07% to 22.92%, with seven samples having RSD_R around 10% or less. The remaining four samples had very low TIA, and their RSD_R values ranged from 13.34% to 22.92%. The study has demonstrated reliable performance of the proposed AOCS method. Several collaborators carried out additional experiments addressing some aspects of the method, leading to further refinements. The proposed method is undergoing evaluation by the AOCS Uniform Methods Committee for adoption as an Official Method for measuring TIA in various legume and grain products.     

Master batch approach for developing PVDF/EVA/CNT nanocomposites with co-continuous morphology and improved electrical conductivity

Conducting polymer composites constituted by co-continuous poly (vinylidene fluoride) (PVDF)/ ethylene- vinyl acetate copolymer (EVA) blends with multiwalled carbon nanotube (CNT) were prepared by melt mixing using different procedures. The effect of the master batch approach on the conductivity, morphology, mechanical, thermal and rheological properties of PVDF/EVA/CNT nanocomposites was compared with that based on one step mixing strategy. The selective extraction experiments revealed that CNT was preferentially localized in the EVA phase in all situations, even when PVDF@CNT master batch was employed. Nanocomposites prepared with EVA@CNT master batch displayed higher conductivity, whose value reached around 10⁻¹ S m⁻¹ with the addition of 0.56 vol% of CNT. The better electrical performance was attributed to the better distribution of the filler, as indicated by transmission electron microscopy and rheological behavior. The electrical and rheological behavior were also investigated as a function of the CNT content.     

Long-term effect of diesel degradation on polyoxymethylene at different temperatures

Diesel is an important fuel, partly because of the longevity and cleanliness of diesel engines. Often, polymers are in direct contact with diesel and understanding compatibility is critical. Polyoxymethylene (POM) is a thermoplastic used to manufacture automotive fuel pump gears and rotors due to its low coefficient of friction and thermal and dimensional stability. In this study, tensile tests were performed on plain and glass fiber reinforced (POM and POMGF) after immersion in diesel at different temperatures (−10°C, 23°C, and 60°C) for 1000, 2000, 3000, 5000, and 10 000 hour. A mathematical model was developed using data from just three tensile stress-strain curves obtained at two different fluid temperatures and three different immersion times. Model and experimental results show good agreement with one another for all conditions tested.     

Thermal dehydrochlorination of pure PVC polymer: Part I—thermal degradation kinetics by thermogravimetric analysis

Thermal degradation of PVC occurs in two stages, with each stage subdivided into two substages. The first refers to the dehydrochlorination, where hydrochloric acid is formed, and giving polyene structures. Hitherto, the degradation mechanism and action of hydrochloric acid as a catalyst during the dehydrochlorination stage are poorly known. Recently, the importance of the tacticity has gained attention for its influence on the dehydrochlorination mechanism. The present work focused on the dehydrochlorination stage, studying the molecular structure by FTIR analysis and the kinetic parameters by TGA analysis in Nitrogen atmosphere, based on three mathematical methods: Friedman, Kissinger, and Flynn-Wall-Ozawa. The sample was a pure homopolymer obtained by suspension polymerization. The dehydrochlorination kinetics follows a first order reaction model and occurs by nucleation and growth. The dehydrochlorination begins with the loss of very labile chlorine atoms present in defective and isotactic molecular segments. The formed HCl acts as a catalyst in the degradation. Following 40% conversion, a drop in Ea is observed. After that, chlorine atoms present in syndiotactic and atactic sequences, are released and, added to the large number of polyene chain sequences, and an increase in Ea is observed up to 60% conversion, where the dehydrochlorination stage is concluded.     

Analysis of surface integrity for minimum quantity lubricant—MQL in grinding

The quality of machined components is currently of high interest, for the market demands mechanical components of increasingly high performance, not only from the standpoint of functionality but also from that of safety. Components produced through operations involving the removal of material display surface irregularities resulting not only from the action of the tool itself, but also from other factors that contribute to their superficial texture. This texture can exert a decisive influence on the application and performance of the machined component. This article analyzes the behavior of the minimum quantity lubricant (MQL) technique and compares it with the conventional cooling method. To this end, an optimized fluid application method was devised using a specially designed nozzle, by the authors, through which a minimum amount of oil is sprayed in a compressed air flow, thus meeting environmental requirements. This paper, therefore, explores and discusses the concept of the MQL in the grinding process. The performance of the MQL technique in the grinding process was evaluated based on an analysis of the surface integrity (roughness, residual stress, microstructure and microhardness). The results presented here are expected to lead to technological and ecological gains in the grinding process using MQL.     

Genetic programming for QoS-aware web service composition and selection

Web services, which can be described as functionality modules invoked over a network as part of a larger application are often used in software development. Instead of occasionally incorporating some of these services in an application, they can be thought of as fundamental building blocks that are combined in a process known as Web service composition. Manually creating compositions from a large number of candidate services is very time consuming, and developing techniques for achieving this objective in an automated manner becomes an active research field. One promising group of techniques encompasses evolutionary computing, which can effectively tackle the large search spaces characteristic of the composition problem. Therefore, this paper proposes the use of genetic programming for Web service composition, investigating three variations to ensure the creation of functionally correct solutions that are also optimised according to their quality of service. A variety of comparisons are carried out between these variations and two particle swarm optimisation approaches, with results showing that there is likely a trade-off between execution time and the quality of solutions when employing genetic programming and particle swarm optimisation. Even though genetic programming has a higher execution time for most datasets, the results indicate that it scales better than particle swarm optimisation.     

Characterizing peers communities and dynamics in a P2P live streaming system

Despite the large number of works devoted to understand P2P live streaming applications, most of them put forth so far rely on characterizing the static view of these systems. In this work, we characterize the SopCast, one of the most important P2P live streaming applications. We focus on its dynamics behavior as well as on the community formation phenomena. Our results show that SopCast presents a low overlay topology diameter and low end-to-end shortest path. In fact, diameter is smaller than 6 hops in almost 90 % of the observation time. More than 96 % of peers’ end-to-end connections present only 3 hops. These values combined may lead to low latencies and a fast streaming diffusion. Second, we show that communities in SopCast are well defined by the streaming data exchange process. Moreover, the SopCast protocol does not group peers according to their Autonomous System. In fact, the probability that a community contains 50 % of its members belonging to the same AS (when we observe the largest AS of our experiments) is lower then 10 %. Peers exchange more data with partners belonging to the same community instead of peers inside the same AS. For the largest AS we have, less than 18 % of peer traffic has been exchanged with another AS partners. Finally, our analysis provides important information to support the future design of more efficient P2P live streaming systems and new protocols that exploit communities’ relationships.     

Spectroscopic Multicomponent Analysis Using Multi-objective Optimization for Variable Selection

The multiple determination tasks of chemical properties are a classical problem in analytical chemistry. The major problem is concerned in to find the best subset of variables that better represents the compounds. These variables are obtained by a spectrophotometer device. This device measures hundreds of correlated variables related with physicocbemical properties and that can be used to estimate the component of interest. The problem is the selection of a subset of informative and uncorrelated variables that help the minimization of prediction error. Classical algorithms select a subset of variables for each compound considered. In this work we propose the use of the SPEA-II （strength Pareto evolutionary algorithm II）. We would like to show that the variable selection algorithm can selected just one subset used for multiple determinations using multiple linear regressions. For the case study is used wheat data obtained by NIR （near-infrared spectroscopy） spectrometry where the objective is the determination of a variable subgroup with information about E protein content （%）, test weight （Kg/HI）, WKT （wheat kernel texture） （%） and farinograph water absorption （%）. The results of traditional techniques of multivariate calibration as the SPA （successive projections algorithm）, PLS （partial least square） and mono-objective genetic algorithm are presents for comparisons. For NIR spectral analysis of protein concentration on wheat, the number of variables selected from 775 spectral variables was reduced for just 10 in the SPEA-II algorithm. The prediction error decreased from 0.2 in the classical methods to 0.09 in proposed approach, a reduction of 37%. The model using variables selected by SPEA-II had better prediction performance than classical algorithms and full-spectrum partial least-squares.     

Flatness-Based Control of a Single Qubit Gate

This work considers the open-loop control problem of steering a two-level quantum system from any initial to any final condition. The model of this system evolves on the state space , having two inputs that correspond to the complex amplitude of a resonant laser field. A symmetry preserving flat output is constructed using a fully geometric construction and quaternion computations. Simulation results of this flatness-based open-loop control are provided.