https://underline.

io/conferences
from October 20 to 23, 2020

Bem-vindos à Nona edição do Brazilian Conference on Intelligent Systems

(BRACIS), ao 16º Encontro Nacional de Inteligência Artificial e Computacional
(ENIAC), e ao 12º Concurso de Teses e Dissertações em Inteligência Artificial e
WELCOME
Computacional (CTDIAC). Neste ano, o BRACIS 2020 está sendo realizado
em conjunto com o VIII Symposium on Knowledge Discovery, Mining and MESSAGE
Learning (KDMiLe) e o IV Brazilian Competition on Knowledge Discovery in
Databases (KDD-BR).

Os cinco eventos combinados tem o objetivo de promover as pesquisas em

Sistemas Inteligentes e o intercâmbio científico entre pesquisadores,
desenvolvedores, cientistas e engenheiros em IA, IC e áreas correlatas.

Nesta edição, devido à pandemia e às questões sanitárias e econômicas

relacionadas, o BRACIS 2020 será uma conferência virtual (online) através da
plataforma da Underline.io. Embora a oportunidade de interagir
pessoalmente tenha sido prejudicada, uma conferência virtual implicou em
custos bastante reduzidos para os organizadores e participantes. Trabalhamos
da melhor forma para oferecer palestras e apresentações sem perder o aspecto
social que sempre foi uma parte central do BRACIS.
Eduardo N. Borges, Graçaliz P. Dimuro, Hélida S. Santos e Leonardo R. Emmendorfer
Universidade Federal do Rio Grande (FURG)
1
• GENERAL
PROGRAMME October 20th, 2020
BRACIS ENIAC CTDIAC KDMiLe
9h-10h ETS 1 M1
10h-11h BTS 1 M1
11h-12h Keynote Speaker: Aline Paes (UFF)
12h-13h Lunch
Opening
13h-14h
cerimony
BTS 2
14h-15h M2 KTS 1
Best papers

15h-16h BTS 3 ETS 2 KTS 2

16h-17h CEIA / CEIC Meeting
2
 BTS1 TUESDAY 10h-11h OCTOBER 20

André Machado,
A new hybridization of evolutionary algorithms, GRASP
Geraldo Mauri, Maria
and set-partitioning formulation for the Capacitated
Claudia Boeres, Rodrigo
Vehicle Routing Problem (CVRP)
Rosa

Applying Dynamic Evolutionary Optimization to the Thiago Lafetá, Gina

Multiobjective Knapsack Problem Oliveira

Backtracking Group Search Optimization: A Hybrid Luciano Pacífico, Teresa

Approach for Automatic Data Clustering Ludermir

Dynamic Software Project Scheduling Problem with PSO

and Dynamic Strategies Based on Memory
Gabriel Fontes da Silva,
Leila Silva, André Britto
BTS 1
BRACIS TECHNICAL
Evaluation of metaheuristics in the optimization of
Victor Costa, Felipe SESSION
Laguerre-Volterra Networks for nonlinear dynamic system
Müller
identification

Cristiano Galafassi,
EvoLogic: Intelligent Tutoring System to Teach Logic Fabiane Galafassi, Eliseo
Reategui, Rosa Viccari

Multi-objective Quadratic Assignment Problem: An Bianca Namie K.

approach using a hyper-heuristic based on the choice Senzaki, Sandra Venske,
function Carolina P. Almeida

On Improving the Efficiency of Majorization-Minorization

Leonardo Emmendorfer
for the Inference of Rank Aggregation Models

3
 BTS2 TUESDAY 14h-15h OCTOBER 20

Cephas A. S. Barreto,
A Distance-weighted Selection of Unlabelled Instances for Arthur Gorgônio, João C.
Self-training and Co-training Semi-supervised Methods Xavier-Júnior, Anne
Canuto

BERTimbau: pretrained BERT models for Brazilian Fábio Souza, Rodrigo

Portuguese Nogueira, Roberto
Lotufo
Lucas Cardoso, Vitor
Cirilo A. Santos, Regiane
BTS 2
Decoding machine learning benchmarks
Kawasaki, Ricardo BRACIS BEST PAPERS
Prudêncio, Ronnie Alves SESSION
Jonathan Carvalho,
Impacts of Multiple Solutions on the Lackadaisical Luciano de Souza,
Quantum Walk Search Algorithm Fernando de Paula Neto,
Tiago Ferreira
Parallel Monte Carlo Tree Search In General Video Game Sandro Rigo, William da
Playing Rosa Fröhlich, Luis
Gustavo S. Centeleghe

Towards a Theory of Hyperntensional Belief Change Marlo Souza

4
 BTS3 TUESDAY 15h-16h OCTOBER 20

A deep learning approach for pulmonary lesion Eduardo Pooch, Carla Becker,
identification in chest radiographs Thatiane Alves

Cristiano Steffens, Silvia Botelho,

A Pipelined Approach to Deal with Image Distortion
Paulo Lilles Drews Jr., Lucas R. V.
in Computer Vision
Messias

Pedro P. Rebouças Filho, Lucas

Fernandes, Francisco H.S. Silva,
A Robust Automatic License Plate Recognition System
Elene F. Ohata, Aldísio Medeiros,
for Embedded Devices
Aloisio Lira, Yuri Lenon, Paulo

BTS 3
Rego

Iago Silva, Gibson Barbosa,

Assessing Deep Learning Models for Human-Robot
Carolina Cani, Assis Filho, Judith
Collaboration Collision Detection in Industrial
Kelner, Djamel Sadok, Silvia Lins, BRACIS TECHNICAL
Environments Ricardo Souza SESSION
Ramásio F. Melo, Gustavo
Diagnosis of Apple Fruit Diseases in the Wild with Lameirão, Guilherme Correia,
Mask R-CNN Bruno Zatt, Marilton Aguiar,
Gilmas Nachtigall, Ricardo Araújo

Geovani Martins, Daniel Ferreira,

Ensemble of algorithms for multifocal cervical image
Fátima Medeiros, Geraldo
segmentation
Ramalho

Improving Face Recognition Accuracy for Brazilian Jones Jose Silva Junior
Faces in a Criminal Investigation Department Anderson da Silva Soares

Neural Architecture Search in Graph Neural Networks Gisele Pappa, Matheus Nunes

5
 Keynote speaker: Aline Paes (UFF)
October 20th 11h

Learning Representations for Natural Language Processing:

methods, challenges, and applications

Language development is considered one of the most significant

turning points in the evolution of human intelligence. Hence,
Short Bio understanding natural language has been one of the grand
Aline Paes is a professor in the Institute of Computing at
Universidade Federal Fluminense (UFF), and a “Young
challenges of Artificial Intelligence since its early days. While we are
Scientist of Our State” by FAPERJ. She leads the research still far from making our computers understand language and
group MeLLL-UFF (Machine Learning and Language
Learning) virtual lab at UFF. She holds a D.Sc. and an capture meaning as we do, recently we have seen tremendous
M.Sc. degree in Systems Engineering and Computer
Science from PESC/COPPE, UFRJ. During the doctorate, advances in several natural language tasks tackled with neural
she was a visiting scholar at Imperial College London, UK.
Aline works in Computer Science, with an emphasis on
network language models. In this talk, we will first visit the early
Artificial Intelligence, with interests and contributions in
the following topics: relational machine learning,
attempts of using Machine Learning to rely on distributional
integrated with neural, statistical and logical techniques, semantics when solving Natural Language Process tasks. We will
natural language processing, updating and adapting
models by transfer learning, theory review, explainable AI, then discuss the recent approaches to learn language models using
induction of programs, games and AI for social good. Aline
Paes has regularly published articles in one of the leading
Deep Learning, including static and contextualized embeddings, and
journals of the ML (Machine Learning Journal), among
others, and national and international conferences of
the plethora of tasks addressed with them. We will see how transfer
Artificial Intelligence. She regularly participates in the learning has enhanced the field’s possibilities in the last couple of
program committee of the major international conference
of Artificial Intelligence, namely AAAI and IJCAI, among years. Finally, we will examine the open challenges that the area still
others, and acts as an ad-hoc reviewer of international
journals. She has experience coordinating research
has with a glance at ethics and responsible AI.
projects approved by funding agencies, including CNPQ
Universal, FAPERJ APQ1, and JCNE, all of them in the area
of Artificial Intelligence and Machine Learning.

6
ETS 1
ENIAC TECHNICAL SESSION EVOLUTIONARY COMPUTATION

ETS 1 TUESDAY 9h-10h OCTOBER 20

A Novel Strategy for Selecting Individuals in Manyobjective Algorithms with Local Search Applied Jorge Nascimento, Danilo Araújo,
to the Network Design Problem Péricles Miranda
A Parallel Strategy for a Genetic Algorithm in Routing Wavelength Assignment Problem Using GPU Esdras La-Roque, Cassio Batista,
with CUDA Josivaldo de Araujo
Image Clustering Based on a Hybrid Group Search Optimization and K-Means Approach for
Luciano Pacífico
Automatic Plant Disease Segmentation
Sérgio Ricardo de Souza, Anolan
Proposta de paralelização em GPUs CUDA do algoritmo MPS para resolução do K-Shortest
Milanés, Alvaro Espíndola, Daniel
Loopless Paths em grafos direcionados e não direcionados
Reis
A C++ Library for Developing Evolutionary Algorithms to the QCaRS Problems Tiago Funk, Fernando Santos
A Multilevel Thresholding Approach Based on Improved Particle Swarm Optimization for Color Larissa Britto, Luciano Pacífico,
Image Segmentation Teresa Ludermir
Volmir Fiorini Júnior, Sandra
Hybrid Algorithm for the Multi-objective Permutation Flow Shop Problem Venske, Carolina Paula de
Almeida
On the Analysis of Mutation Operators in Multiobjective Cartesian Genetic Programming for
Lucas Souza, Heder Bernardino
Designing Combinational Logic Circuits
Gustavo Pacheco, Eduardo
Using Genetic Algorithms to Design an Optimized Keyboard Layout for Brazilian Portuguese
Palmeira, Keiji Yamanaka

7
7
ETS 2
ENIAC TECHNICAL SESSION TEXT AND WEB MINING

ETS 2 TUESDAY 15h-16h OCTOBER 20

A Framework for Multi-Document Extractive Summarization of Reviews with Aspect-Based André Seidel Oliveira, Anna
Sentiment Analysis Costa, Eduardo Hruschka
A Cooking Recipe Multi-Label Classification Approach for Food Restriction Identification Larissa Britto, Luciano Pacífico,
Emilia Oliveira, Teresa Ludermir
A glance of gastronomic tourism: A case on TripAdvisor
Fabio Lobato, Jorge Silva Junior,
Luiz Carlos Fernandes Junior

A Sentiment Classification Approach for Books Reviews in Brazilian Portuguese Using Different
Larissa Britto, Luciano Pacífico
Feature Extraction Methods
Automatic Cooking Recipe Difficulty Level Inference using Natural Language Processing Techniques Larissa Britto, Luciano Pacífico,
Teresa Ludimir
Caio Mota, Andressa Lima, André
Classification of Court Lawsuits Pages using Multimodal Convolution Neural Networks
Nascimento, Péricles Miranda
Rafael Ferreira L. Mello

The Construction of a Corpus for Detecting Irony and Sarcasm in Portuguese Gabriel Schubert

8
8
CTDIAC M.Sc.

M1 TUESDAY 9h-11h OCTOBER 20

Rubens Moraes Filho, Levi Lelis

Asymmetric Action Abstractions for Real-Time Planning in Extensive-Form Games
(Universidade Federal de Viçosa)
Automatic Algorithm Selection for the Quadratic Assignment Problem Using Meta-learning and Augusto Dantas, Aurora Pozo
Fitness Landscape Measures (UFPR)
Marciele Bittencourt, Renato
ML-MDLText: um método de classificação de textos multirrótulo de aprendizado incremental
Silva, Tiago Almeida (UFSCar)

Semi-Supervised Self-Organizing Maps with Time-Varying Structures for Clustering and Pedro Braga, Hansenclever F.
Classification Bassani (UFPE)

M2 TUESDAY 14h-15h OCTOBER 20

Wesley Kaizer, André Grahl

Sequencing Operator Counts with State-Space Search
Pereira, Marcus Ritt (UFRGS)
Rodrigo Azevedo Santos, Gerson
Transfer Learning by Mapping and Revising Boosted Relational Dependency Networks Zaverucha (UFRJ), Aline M. Paes
Carvalho (UFF)

9
9
KTS 1
KDMILE TECHNICAL SESSION

KTS 1 APPLICATIONS PART 1 TUESDAY 14h-14h30 OCTOBER 20

Improving automatic data extraction from financial statements with clustering Victor Ferraz, Gabriel Olivato, Igor
analysis Magollo, Murilo Naldi
Luiz Felipe Vercosa, Rodrigo Lira,
Rodrigo Monteiro, Kleber Silva,
Impact of Unusual Features in Credit Scoring Problem
Jailson Magalhaes, Alexandre Maciel,
Byron Leite, Carmelo Bastos-Filho

Marcelo de Caux, Flavia Bernardini,

Short-term Forecasting in Bitcoin Timeseries Using LSTM and GRU RNNs
Jose Viterbo

Daniela Giacomelli, Murilo Naldi,

Acidentes de trabalho no Brasil: uma análise descritiva
Elaine Faria

KTS 1 APPLICATIONS PART 2 TUESDAY 14h30-15h OCTOBER 20

Evaluation of the Usefulness of Explanations of Post-hoc Interpretability for Vinícius Araújo

Malaria Detection Leandro Marinho

Machine Learning to Assist in Pneumonia Decision Making: A Systematic Review Victor Silva, Amanda Days,
of the Literature Damires Souza, Alex Rêgo

Juliana B. Mattos, Renato Vimieiro, Paulo

Clinical risk factors of ICU & fatal COVID-19 cases in Brazil
S.G. de Mattos Neto, Eraylson G. Silva

Towards ideal time window for classifying motor imagery in brain-computer Vitor Mendes Vilas-Boas, Vitor da Silva
interfaces Jorge, Cleison Daniel Silva.

10
10
KTS 2
KDMILE TECHNICAL SESSION

KTS 2 MACHINE LEARNING AND NATURAL

TUESDAY 15h-15h30 OCTOBER 20
LANGUAGE PROCESSING
P. Neto Berlanga, E. E. S. Ruiz,
Experimenting split-and-rephrasing sentences using part-of-speech labels
E. Y. Okano

CONLL Dependency Parser: Extrinsic Evaluation through the Open Information

Jardel Baia, Arley Prates, Daniela Claro
Extraction task

From audio to information: Learning topics from audio transcripts João Rodrigues, Emerson Paraiso

Renato Geh, Denis Mauá,

Learning Probabilistic Sentential Decision Diagrams by Sampling
Alessandro Antonucci

KTS 2 MACHINE LEARNING IN

TUESDAY 15h30-16h OCTOBER 20
BRAZILIAN ELECTIONS

Leandro G. C. Simoes, Filipe A. N.

Spending Segmentation and Outlier Detection in Brazilian Elections
Verri, Takashi Yoneyama

André Cristiani, Douglas Lieira,

A Sentiment Analysis of Brazilian Elections Tweets
Heloisa Camargo

Lucas Henrique M. Jacintho, Tiago P.

Brazilian Presidential Elections: Analyzing Voting Patterns in Time and Space Using a
Silva, Antonio Rafael S. Parmezan,
Simple Data Science Pipeline
Gustavo Enrique A. P. Alves Batista

11
11
• GENERAL October 21st, 2020
PROGRAMME
BRACIS ENIAC CTDIAC KDMiLe
9h-10h ETS 3
10h-11h BTS 4
11h-12h Keynote Speaker: André Paim (Loggi)
12h-13h Lunch
13h-14h D1
14h-15h BTS 5 D1 KTS 3
ETS 4 KTS 4
15h-16h BTS 6
Best papers Best papers

16h-17h Industry Panel 1

17h-18h CEIA Meeting
12
 BTS4 WEDNESDAY 10h-11h OCTOBER 21

A Multi-Level Approach to the Formal Semantics of Alison Panisson, Rafael Bordini,

Agent Societies Antonio Carlos da Rocha Costa

An Argumentation-based Approach for Explaining Mariela Morveli-Espinoza,

Goals Selection in Intelligent Agents Cesar Tacla, Henrique Jasinski

Application-Level Load Balancing for Reactive

Igor Pereira, Lisane Brisolara,
Wireless Sensor Networks: An Approach Based on
Paulo Roberto Ferreira Jr
Constraint Optimization Problems
Leonardo F. Costa, Matheus

BTS 4
Araújo, Vinicius Sampaio,
Cooperative Observation of Smart Target Agents Thayanne da Silva, João Andrade,
Raimundo Ferro Junior, Gustavo
Campos, Gabriel Melo BRACIS TECHNICAL
Milton Condori, Leliane N. SESSION
Finding Feasible Policies for Extreme Risk-Averse
Barros, Karina Valdivia-Delgado,
Agents in Probabilistic Planning
Valdinei Freire, Denis Mauá

On the Performance of Planning through Leliane N. Barros, Denis Mauá,

Backpropagation Thiago Bueno, Renato Scaroni

Karina Valdivia-Delgado, Elthon

Risk Sensitive Stochastic Shortest Path and
Manhas de Freitas, Valdinei
LogSumExp: from theory to practice
Freire

Testing Multiagent Systems under Organizational

Ricardo Arend, Eder Mateus
Model Moise using a Test Adequacy Criterion
Gonçalves
based on State Transition Path

13
 BTS5 WEDNESDAY 14h-15h OCTOBER 21

A Differential Evolution Algorithm for Contrast Artur Leandro da Costa

Optimization Oliveira, André Britto

Frederico Möller, Heder

A Reinforcement Learning Based Adaptive Mutation
Bernardino, Luciana B.
for Cartesian Genetic Programming Applied to the
Gonçalves, Stênio Sã Rosário
Design of Combinational Logic Circuits
Soares

An Evolutionary Algorithm for Learning Interpretable Henry Cagnini, Alex Freitas,

Ensembles of Classifiers Rodrigo Barros

BTS 5
Renan Goulart, Saulo Villela,
An Evolutionary Analytic Center Classifier Carlos C. Borges, Raul
Fonseca
BRACIS TECHNICAL
Matheus G. Pires,
Genetic Learning Analysis of Fuzzy Rule-Based
Fabiana Bertoni, Allen
SESSION
Classification Systems considering Data Reduction
Hichard M. Santos

Improving FIFA Player Agents Decision-Making

Matheus Faria, Rita Maria
Architectures based on Convolutional Neural
Silva Julia, Lídia Tomaz
Networks through Evolutionary Techniques

On the Multiple Possible Adaptive Mechanisms of the Victor Costa,

Continuous Ant Colony Optimization Felipe Müller

Ana Carolina L.C. Queiroz,

Solving Multi-Agent Pickup and Delivery Problems
Heder Bernardino, Alex
Using a Genetic Algorithm
Borges Vieira, Helio Barbosa

14
 BTS6 WEDNESDAY 15h-16h OCTOBER 21

A computational tool for automated detection of Eduardo Pooch, Carla

genetic syndrome using facial images Becker, Thatiane Alves

Ruan C. Rodrigues, Marcelo

Domain Adaptation of Transformers for English Word Inuzuka, Hugo do
Segmentation Nascimento, Acquila
Santos Rocha

Impact of Text Specificity and Size on Word Thiago Dal Pont, Isabela
Embeddings Performance: an Empirical Evaluation in Sabo, Jomi Hübner, Aires J
Brazilian Legal Domain Rover

Henrique Tavares,
BTS 6
People Identification Based on Soft Biometrics João B. Cardia, Joao Papa, BRACIS TECHNICAL
Features Obtained from 2D Poses Danilo Colombo, Aparecido SESSION
Marana

Everton Santana, Ricardo

Photovoltaic Generation Forecast: model training and
Petri, Bruno Zarpelão,
adversarial attack aspects
Sylvio Barbon Junior

Elias Paulino Medeiros,

Texture analysis based on structural co-occurrence
Geraldo Ramalho, Daniel
matrix improves the colorectal tissue characterization
Ferreira

Claudio Mello Jr., Lucas

Unsupervised Learning Method for Encoder-Decoder- Ricardo Vieira Messias,
Based Image Restoration Paulo Lilles Drews Jr, Silvia
Botelho

15
 Keynote speaker: André Paim (Loggi)
October 21st 11h

Loggi - Data Science Applications for problem solving in the

Logistics Chain

Brazilian logistics sector still has great challenges to be overcome

when compared to those of large world economies. This lecture will
give an overview of the objectives, activities and solutions of Loggi
Short Bio
André Paim Lemos has a degree in Computer Science from Tecnologia Ltda, established in 2013. Loggi has positioned itself, in an
the Federal University of Minas Gerais (UFMG) in 2003
and a master's degree (2007) and a doctorate (2011) in
unprecedented way, to be an express logistics solution with national
Electrical Engineering with an emphasis on Artificial coverage, facilitating the growth of a new trade in Brazil, with
Intelligence from the same university. Between 2011 and
2018 worked as a full professor in the Department of economy, agility and reliability. By 2021, Loggi's goal is to connect all
Electronic Engineering at UFMG. During this period he
taught undergraduate and graduate courses in the areas of Brazilians, covering 100% of the cities, a goal that demands high
Automation, Distributed Systems, Real
Programming, Computational Intelligence and Machine
Time quality development processes and methods and human resources.
Learning; supervised several undergraduate, master's and The development and evolution of software solutions at Loggi is
doctoral students; published more than 70 articles in
journals and conference proceedings in the field of carried out based on evidence, from historical data analyzes using
Artificial Intelligence; and participated in R&D projects
related to Data Mining and Artificial Intelligence data science methods and techniques. Some examples of application
applications in industrial problems financed by companies
such as Petrobras, CEMIG, Gerdau, CHESF, among others.
of data sciences in Loggi solutions in the logistics chain will be
He currently works as Head of Data Analytics at Loggi, illustrated, such as routing, allocation of routes to couriers, package
where he coordinates data management and analysis
projects in the areas of Business Intelligence, Data tracking, among others. To learn more about Loggi, visit:
Engineering and Data Science. Currently he also serves as
a Collaborating Professor in the Graduate Program at the loggi.com/venha.
Faculty of Electrical and Computer Engineering (FEEC) at
the University of Campinas (Unicamp).

16
ETS 3
ENIAC TECHNICAL SESSION MACHINE LEARNING I

ETS 3 WEDNESDAY 9h-10h OCTOBER 21

Igor Ventorim, Diego Luchi, Flávio

A biased sampling method for applying DBSCAN
Miguel Varejão
Felipe Farias, Teresa Ludermir,
Clustering for Data-driven Unraveling Artificial Neural Networks
Carmelo Bastos-Filho
Douglas Meneghetti, Reinaldo
Towards Heterogeneous Multi-Agent Reinforcement Learning with Graph Neural Networks
Bianchi
Samuel Henrique Silva, Fabricio
Use of Convolutional Neural Networks to Identify Focal Cortical Dysplasia in Patients with
Simozo, Luiz Otavio Murta Junior,
Refractory Epilepsy
Renato Tinos
Ana Paula Sodré, Luis E. Floriano,
A Clustering Visualization Query Language
Aurora Pozo, Carmem Hara
Aprendizado Profundo Aplicado na Previsão de Receita Tributária Utilizando Variáveis
Karla Figueiredo, Priscila Silva
Endógenas
Larissa Britto, Luciano Pacífico,
Medicinal Plant Recognition Using Color, Texture and Shape Features Matheus Fidelis da Silva, Teresa
Ludermir
Rafael Silva, Myriam Delgado,
Training Data Filtering for Deep Learning Applied to Inspection of Welded Joints in Oil Pipelines
Ricardo D. Silva, Fernando Suyama

17
ETS 4
ENIAC TECHNICAL SESSION RUNNERS UP BEST PAPER

ETS 4 Main Track WEDNESDAY 15h-16h OCTOBER 21

Pablo Jaskowiak, Angelo Pelisson,

Comparative Study of Photovoltaic Power Forecasting Methods
Thiago Covoes, Anderson W. Spengler
Gated Recurrent Unit Networks and WaveletsDiscrete Transformations Applied to
Victor Biazon, Reinaldo Bianchi
Forecastingand Trading in the Stock Market

Caio Ponte, João José Vasco Furtado,

Optimizing Random Forest from the pondering of regression tree leaves
Carlos Caminha Neto

Sample Bias Effect on Meta-Learning Mariane Reis, Ana Carolina Lorena

Similarity Search using the NK Interaction Graph José Carlos B. Moraes, Renato Tinos

Undergraduate track
Avaliação empírica de classificadores e métodos de balanceamento para detecção de Victor Nicola, Marcelo Lauretto, Karina
fraudes em transações com cartões de créditos Valdivia-Delgado

Francisco Neto, Romero Silva, Roberta

Cloud Computing and Machine Learning for Analysis of Large Volumes of Educational Data
Gouveia, Maria Batista, Igor Gomes de
Networks
Oliveira

Lucas Dias, Péricles Miranda, André

Nascimento, Filipe Cordeiro, Rafael F.L.
Meta-Characteristics Extraction from Image Datasets for Selection of Convolutional Neural
Mello, Ricardo Prudêncio, Ricardo
Oliveira

Unsupervised Machine Learning Based on Heterogeneous Networks for Text Clustering José Vitor Santos, Rafael Rossi

Using Artificial Neural Networks to Classify Treadmill Running Patterns Sergio Baldo, Paulo Santiago,
in High-Performance Sports Renato Tinos 18
CTDIAC Ph.D.

D1 WEDNESDAY 13h30-15h OCTOBER 21

Fernando de Paula Neto (UFPE),

Avanços em Redes Neurais Quânticas Teresa Ludermir (UFPE), Wilson
de Oliveira (UFRPE)
Ramon Fraga Pereira, Felipe
Goal Recognition over Imperfect Domain Models
Meneguzzi (PUCRS)
Felipe Leno da Silva (Advanced
Methods and Algorithms for Knowledge Reuse in Multiagent Reinforcement Learning Institute for AI),
Anna Costa (USP)

19
19
KTS 3
KDMILE TECHNICAL SESSION

KTS 3 NEURAL NETWORKS AND REINFORCEMENT

WEDNESDAY 14h-14h30 OCTOBER 21
LEARNING
Laercio Oliveira Junior,
Clustered Echo State Networks for Signal Observation and Frequency Filtering
Florian Stelzer, Liang Zhao

Inducing selfish agents towards social efficient solutions João Schapke, Ana Bazzan

Andrey de Aguiar Salvi,

An Experimental Analysis of Model Compression Techniques for Object Detection
Rodrigo Coelho Barros

Guilherme Santos,
Accelerating learning of route choices with C2I: a preliminary investigation
Ana Bazzan

KTS 3 TEXT MINING: PART 1 WEDNESDAY 14h30-15h OCTOBER 21

Breno Cardoso, Denilson

Evaluating an Aspect Extraction Method for Opinion Mining in the Portuguese Language
Pereira
Eliseu Guimarães,
Transfer learning for Twitter sentiment analysis: Choosing an effective source dataset Jonnathan Carvalho, Aline
Paes, Alexandre Plastino

Combining compact news representations generated using DistilBERT Carlos Abel Córdova Sáenz,
and topological features to classify fake news Marcelo Dias, Karin Becker

Marcelo Inuzuka, Hugo

Nascimento, Fernando
Doclass: opensource software to support document labeling and classification
Almeida, Bruno Barros,
Walid Jradi

20
20
KTS 4
KDMILE TECHNICAL SESSION

KTS 4 TEXT MINING: PART 2 WEDNESDAY 15h-15h30 OCTOBER 21

Pedro V. Brum, Matheus C. Teixeira,

A Characterization of Portuguese Tweets Regarding the Covid-19 Pandemic Renato Vimieiro, Wagner Meira
Jr, Gisele L. Pappa, Renato Miranda
Carlos Euzebio, Sidney Agy, Boldorini Jr.
Statistical analysis of small twitter data collection to identify dengue outbreaks Claudio, José Renato Alcarás, Lucas
Porto, Alexandre Martinez, Evandro Ruiz

Forecasting future corn and soybean prices: an analysis of the use of textual Ivan José dos Reis Filho, Guilherme B.
information to enrich time series Correa, Guilherme Mendonça Freire,
Solange Oliveira Rezende.

KTS 4 BEST PAPERS WEDNESDAY 15h30-16h OCTOBER 21

Marcos Wander Rodrigues, Luis

Extreme Events Characterization on Time Series
Enrique Zárate

Prediction of Environmental Conditions for MaritimeNavigation using a Network of Caio Netto, Denis Mauá, Eduardo
Sensors: A Practical Application of Graph Neural Networks Tannuri, Fábio Cozman

Quarenteners vs. Cloroquiners: a framework to analyze the effect of political Régis Ebeling, Carlos Córdova,
polarization on social distance stances Jeferson Campos Nobre, Karin Becker

João Antonio O. Pedrosa, Derick

Automated classification of cardiology diagnoses in textual medical reports
Oliveira, Wagner Meira Jr., Antônio
Ribeiro
21
21
 Sergio Novaes - Advanced Institute for Artificial Intelligence
He is Full Professor of Physics at the São Paulo State University (Unesp), Scientific Director of the
Center for Scientific Computing from Unesp and the co-founder of the Advanced Institute for Artificial
Intelligence (AI2). He obtained the B.Sc. and Ph.D. degrees in Physics from the University of São Paulo
(USP) and he was a Postdoctoral Research Fellow at the Lawrence Berkeley National Laboratory
(Berkeley, USA). He was visiting researcher at the University of Wisconsin (Madison, USA), University of
Valencia (Spain), and at the Fermi National Accelerator Laboratory (Chicago, USA). He has been the PI

Industry panel 1 of several R&D projects associated to the private sector (Padtec, Intel, Huawei, etc.) which include an
Intel Parallel Computing Center (IPCC) and a Center of Excellence in Machine Learning, and Huawei is
supporting the development of Kytos, a new SDN Controller. He and his team have a partnership for

October 21st 16h

more than a decade with Caltech which set the record of data transmission between the North and
South Hemispheres three times during the bandwidth challenges at the SuperComputing conferences.

Davi Reis - Loggi

Davi Reis, Computer Scientist, started his career working as a developer and white hat hacker in the
world of ISPs. During his MSc, working as a database and information retrieval scientist, created the
RTDM algorithm, co-authored the CMPH open source library and joined Akwan, the startup that
became Google Brazil. He worked ten years at Google, leading projects in Search, Mobile and Social
Ads. He brought Orkut from zero revenue to profitability. He also led project Google Pigeon, improving
local search for billions of people. Co-founded, almost scaled and acquisold WorldSense. He is
currently helping rebuild Brazil's logistics as the CTO for Loggi.

Hans De Canck - AI Experience Centre

Hans De Canck manages and directs the development of the AI Experience Center of the Vrije
Universiteit Brussels (VUB). The AI Experience Center is a state of the art Digital Innovation Hub in
Brussels. Several research centers from the Vrije Universiteit Brussel have joined forces to develop a
multi-disciplinary Research and Innovation offering on AI towards Academia, Industry, Policy makers
and the broader public. The AI Experience Center will accelerate this offering and collaborate with
other stakeholders in the open innovation ecosystem. Hans works with the research teams across the
university and acts as a coordinator for the AI for the Common Good initiative, launched early 2019. He

Applied AI: has a background in research and IT development. The last 15 years he held different management
positions in Research and Technology Organisations like iMinds, imec and VITO.

Bridging the Gap

Between Academy Stênio Fernandes - ElementAi

and Industry
Stenio Fernandes is one of the leaders in the AI Platform group at Element AI in Montreal, Canada. He
is leading a team of Applied Research Scientists and AI Developers in the context of Time-dependent
problems, such as Time Series Forecasting, Anomaly Detection, Spatio-Temporal modelling, and
Concept Drifting. Element AI (EAI) models have been applied to several problems to different industry
verticals, such as Capital Markets, Retail, Cybersecurity, Manufacturing, Insurance, and Transportation
and Logistics. Currently, Stenio's team is fine-tuning EAI's state-of-the-art models on time series
forecasting and making them viable to integrate into EAI products and services' portfolio. He provides
technical, strategic, and intellectual leadership to all team members. He holds a Ph.D. in Computer
Science (UFPE) with Post-Doctoral experience conducted at the University of Ottawa, Canada. He is an
IEEE Senior Member and a certified Project Management Professional - PMP. As a former Professor of
Computing Science, he was involved in dozens of scientific research projects in Brazil (UFPE) and
Canada (University of Ottawa and Carleton University). He published more than 140 scientific research
papers in major peer-reviewed conferences and journals. His citation indexes are: h-index = 18, i10-
index = 30.
22
• GENERAL October 22nd, 2020
PROGRAMME
BRACIS ENIAC KDD-BR
9h-10h ETS 5
10h-11h BTS 7
Keynote Speaker: Jose M. Alonso
11h-12h
(CiTIUS, Univ. de Santiago de Compostela, Spain)
12h-14h Lunch
14h-15h BTS 8
15h-16h BTS 9 ETS 6
16h-17h C3 Live - FURG Top three teams

17h-18h C4AI Live

23
 BTS7 THURSDAY 10h-11h OCTOBER 22

A Study on the Impact of Intradomain Finetuning Luiz H. Bonifacio, Paulo

of Deep Language Models for Legal Named Entity Augusto A. Vilela, Gustavo
Recognition in Portuguese Lobato, Eraldo Fernandes

Pre-trained Data Augmentation for Text Hugo Queiroz Abonizio,

Classification Sylvio Barbon Junior

Arthur Reys, Danilo Silva,

Predicting Multiple ICD-10 Codes from Brazilian- Daniel Severo, Saulo Pedro,
Portuguese Clinical Notes Marcia Sá, Guilherme Augusto
Salgado
BTS 7
Semi-Supervised Sentiment Analysis of Portuguese
Tweets with Random Walk in Feature Sample Pedro Lourenço, Filipe Verri
BRACIS TECHNICAL
Networks SESSION
The use of machine learning in the classification of
electronic lawsuits: an application in the Court of Adriano Silva, Luiz Maia
Justice of Minas Gerais

Towards a Free, Forced Phonetic Aligner for Ana Larissa Dias, Cassio Batista,
Brazilian Portuguese Using Kaldi Tools Daniel Santana, Nelson Neto

Twitter Moral Stance Classification using Long Matheus Pavan, Wesley

Short-Term Memory Networks Santos, Ivandré Paraboni

24
 BTS8 THURSDAY 14h-15h OCTOBER 22

A Fuzzy Approach For Classification and Novelty

André Cristiani, Tiago Pinho
Detection in Data Streams Under Intermediate
da Silva, Heloisa Camargo
Latency
Giancarlo Lucca, Eduardo
A Fuzzy Reasoning Method based on ensembles of Borges, Graçaliz Dimuro,
generalizations of the Choquet integral Helida Santos, Tiago Asmus,
José A Sanz, Humberto Bustince

A Useful Tool to Support the Ontology Alignment Miriam dos Santos, Carlos E.
Repair Mello, Tadeu Classe

Aggregation with Weak, Axiological and Strong

Sufficientarian Functions
Henrique Oliveira, João
Fernando Lima Alcântara
BTS 8
BRACIS TECHNICAL
Frederico Tiggemann, SESSION
Bryan S. Pernambuco,
An alternative to Power Measure for Fuzzy Rule- Giancarlo Lucca, Eduardo
Based Classification Systems Borges, Helida Santos,
Gracaliz Dimuro, José Antonio
Sanz, Humberto Bustince

Juliana Mattos, Renato

Exceptional Survival Model Mining Vimieiro, Paulo Mattos Neto,
Eraylson Galdino

Elidiane Nascimento,
FT-BlinGui: a fuzzy-based wearable device system to
Ricardo Rios, Tatiane
avoid visually impaired collision in real time
Nogueira

25
 BTS9 THURSDAY 15h-16h OCTOBER 22

2CS: Correlation-guided Split Candidate Selection in Saulo M. Mastelini, André

Hoeffding Tree Regressors Ponce L. F. de Carvalho

Vinicius Martins, Victor

Active Learning embedded in incremental decision
Turrisi da Costa, Sylvio
trees
Barbon Junior

Adriano Barreto, Igor

Matheus Moreira,
An Online Pyramidal Embedding Technique for High
Claudomiro Souza Junior,
Dimensional Big Data Visualization
Caio Flexa, Eduardo
Cardoso BTS 9
Comparative study of Fast Stacking Ensembles families
Laura P. Mariño, Agustín A. BRACIS TECHNICAL
algorithms
Ortiz Díaz, Germano SESSION
Vasconcelos

Ensemble of Binary Classifiers Combined Using Rodolfo Anibal Lobo,

Recurrent Correlation Associative Memories Marcos Eduardo Valle

Link prediction in social networks: an edge creation Argus Barbosa Cavalcante,

history retrieval-based method that combines Claudia Justel, Ronaldo
topological and contextual data Goldschmidt

Particle Competition for Unbalanced Community

Luan Martins, Liang Zhao
Detection in Complex Networks

Towards an Instance-Level Meta-Learning-Based Diego Furtado Silva, Caio

Ensemble for Time Series Classification Ueno, Igor Braga

26
 Keynote speaker: Jose Alonso
(CiTIUS, Univ. de Santiago de Compostela, Spain)
October 22nd 11h

Paving the way from Fuzzy Logic towards Explainable Artificial Intelligence
In the era of the Internet of Things and Big Data, data scientists are required
to extract valuable knowledge from the given data. They first analyze, cure
and pre-process data. Then, they apply Artificial Intelligence (AI) techniques to
automatically extract valuable knowledge from data. Explainable AI (or just
XAI for short) emerges as an endeavor to evolve AI methodologies and
technology by focusing on the development of agents capable of both
generating decisions that a human could understand in a given context, and
Short Bio explicitly explaining such decisions. This way, it is possible to verify if
Dr. Jose M. Alonso received his M.Sc. and Ph.D. degrees in automated decisions are made on the basis of accepted rules and principles,
Telecommunication Engineering, both from the Technical so that decisions can be trusted, and their impact justified in terms of safety,
University of Madrid (UPM), Spain, in 2003 and 2007,
respectively. He is currently "Ramón y Cajal" researcher fairness, robustness, accountability and lineage.
funded by the Spanish Government under project RYC- The main goal of this talk is to provide audience with a holistic view of
2016-19802, affiliated to CiTIUS-USC, secretary of the ACL
Special Interest Group on Natural Language Generation
fundamentals and current research trends in the XAI field. We will pay special
(SIGGEN), board member of the European Society for attention to fuzzy-grounded knowledge representation and reasoning. Fuzzy
Fuzzy Logic and Technology (EUSFLAT), Associate Editor rules relate fuzzy sets and make it feasible to infer meaningful information
of the IEEE Computational Intelligence Magazine granules at certain level of abstraction. Fuzzy modeling favors fairness,
(ISSN:1556-603X), member of the Editorial Board of the
International Journal of Computational Intelligence
accountability, transparency, trustfulness and explainability. Interpretable
Systems (ISSN: 1875-6883), Chair of the IEEE-CIS Task fuzzy models represent knowledge in a way close to natural language, easy to
Force on Explainable Fuzzy Systems, member of the IEEE- interpret and understand by users no matter their background, because such
CIS Task Force on Explainable Machine Learning, models are endowed with linguistic interpretability and global semantics.
member of the IEEE-CIS Task Force on Fuzzy Systems
Software, member of the IEEE-CIS Content Curation Explainable fuzzy systems wrap interpretable fuzzy models with an interactive
Subcommittee. In addition, he is the President of the linguistic interface that makes them self-explanatory. Moreover, explainable
Executive Board and Deputy Coordinator of the H2020- fuzzy systems enhance human-machine interaction through factual and
MSCA-ITN-2019 (Grant Agreement No 860621) project counterfactual multi-modal effective explanations supported by Fuzzy Logic
entitled “Interactive Natural Language Technology for
Explainable Artificial Intelligence” (NL4XAI). He has and interactive Natural Language Technology.
published more than 140 papers in international journals,
book chapters and conferences. His research interests
include explainable artificial intelligence, computational
intelligence, interpretable fuzzy systems, natural language
generation, development of free software tools, etc.

27
ETS 5
ENIAC TECHNICAL SESSION MACHINE LEARNING II

ETS 5 THURSDAY 9h-10h OCTOBER 22

A Network-Based High-Level Data Classification Algorithm Using Betweenness Centrality Esteban Vilca Zuñiga, Liang Zhao

Classifying the Macronutrient Deficiency in Soybean Leaf with Deep Learning Maicon Sartin

Evaluation of Texture Maps as Input to Extract Deep Features in Glaucoma Diagnosis Daniel V. Silva, Romuere Silva

Ronnypetson Souza da Silva, Valter

Inter-affection of Multiple Datasets in Speech EmotionRecognition with Neural Networks
Akira Miasato Filho
Davi Nascimento, Anna Costa
Stock Trading Classifier with Multichannel Convolutional Neural Network
Reinaldo Bianchi
Nathan Garcia, Eduardo Borges,
Death Registry Prediction in Brazilian Male Prisons with a Random Forest Ensemble Giancarlo Lucca, Helida Santos,
Gracaliz Dimuro
Descoberta de Conhecimento em Dados de Scout do Campeonato Brasileiro de Futebol Luís Ortolan, Diego Furtado Silva
Hybrid Approach for Detecting Brazilian Real Coins with Localization Algorithms and
David Yonekura, Elloá B. Guedes
Convolutional Neural Networks

28
ETS 6
ENIAC TECHNICAL SESSION AUTOMATED PLANNING

ETS 6 THURSDAY 15h-16h OCTOBER 22

Mariane Amaral, Anderson Silveira,

A short-term electricity demand forecasting in the Southern Region of Brazil using the ARIMA
Viviane L. D. Mattos, Eduardo Borges
Model and the Holt Exponential Smoothing Model (SEH)

Luiz Otavio Murta Junior, Fabricio

Brain Tissue Classification to Detect Focal Cortical Dysplasia in Magnetic Resonance Imaging
Simozo, Marcos Soares de Oliveiras
Raquel Candido, Rafael Del Lama,
Natália Chiari, Marcello Nogueira-
Impact of Feature Selection on Clustering Images of Vertebral Compression Fractures
Barbosa, Paulo de Azevedo-Marques,
Renato Tinos
Gabriel Crispino, Karina Valdivia-
Monte Carlo Tree Search Algorithm for SSPs Under the GUBS Criterion
Delgado, Valdinei Freire
Henrique Pastor, Valdinei Freire,
Políticas Sensíveis ao Risco para o Controle da Propagação de Doenças Infecciosas Leliane Nunes de Barros, Karina
Valdivia-Delgado

Sensor Validation for Indoor Air Quality using Machine Learning Vagner Seibert, Ricardo Araújo

Guilherme Yambanis Thomaz, Denis

A Contact Network-Based Approach for Online Planning of Containment Measures for COVID-19
Mauá, Leliane Nunes de Barros

29
KDD-BR
TOP THREE TEAMS

• Teams invited to present their solutions

• The top three teams are in alphabetic order and in

this session the actual places will be disclosed.

KDD-BR THURSDAY 16h-17h OCTOBER 22

Adriano Avelar

David Yonekura

João Pedro Peinado

30
 Essa live ocorrerá no contexto do projeto C3 Live, que consiste

C3 Live em uma live semanal, organizada pelo C3/FURG, para discutir

temas diversos com a comunidade do C3 e geral. A ideia é uma
conversa informal sobre o impacto da IA no futuro da
October 22nd 16h humanidade sob diferentes aspectos.

Artur H. Barcelos
Bacharelado em Arqueologia Nelson Duarte Filho
Professor do Instituto de Ciências Professor C3/FURG
Humanas e da Informação ICHI/FURG

Graçaliz P. Dimuro Silvia Botelho

A era da Professora C3/FURG Diretora C3 - Proferssora FURG

Inteligência
Artificial:
Ressignificando
Robôs e Humanos?

31
 October 23rd, 2020
• GENERAL BRACIS ENIAC
PROGRAMME 9h-10h BTS 10 ETS 7
10h-11h Keynote Speaker: Celine Vens
(KULAK, Belgium)

11h-12h Keynote Speaker: Yaochu Jin

(University of Surrey, Guildford, U.K.)

12h-13h Lunch
13h-14h BTS 11
14h-15h BTS 12 ETS 8
15h-16h Industry Panel 2
16h-17h Awards/closing session

32
 BTS10 FRIDAY 9h-10h OCTOBER 23

Jesse Read, Ricardo Rios,

Data Streams are Time Series: Challenging
Tatiane Nogueira,
Assumptions
Rodrigo Mello

Sandro Rigo, William R.

Evaluating a new approach to data fusion in wearable Fröhlich, Clarissa Rodrigues,
physiological sensors for stress monitoring Amanda Jabroski, Andréia
Rodrigues, Elisa Kern Castro

Financial time series forecasting via CEEMDAN-LSTM Renan Avila, Glauber De

with exogenous features Bona

Bruna Guterres, Je Nam

BTS 10
Intelligent Classifiers on the Construction of Pollution Junior, Amanda Gurreiro, BRACIS TECHNICAL
Biosensors Based on Bivalves Behavior Viviane Fonseca, Silvia SESSION
Botelho, Juliana Sandrini

Ana Carolina Lorena, José

Measuring instance hardness using data complexity
Luis Arruda, Ricardo
measures
Prudêncio

Victor Barella, Luís Garcia,

Simulating Complexity Measures on Imbalanced
André Ponce L. F. de
Datasets
Carvalho

SSL-C4.5: Implementation of a classification algorithm Agustín A. Ortiz Díaz, Flávio

for semi-supervised learning based on C4.5 Bayer, Fabiano Baldo

33
 BTS11 FRIDAY 13h-14h OCTOBER 23

AgentDevLaw: A Middleware Architecture for Fábio Aiub Sperotto,

Integrating Legal Ontologies and Multi-Agent Systems Marilton Aguiar

Improved Multilevel Algorithm to Detect Camila Tautenhain, Calvin

Communities in Flight Networks Costa, Mariá Nascimento

Clóvis Daniel S. Silva,

KNN applied to PDG for source code similarity
Leonardo Rocha, Gerardo
classification
Valdisio R. Viana, Leonardo
Ferreira da Costa

New Fast Morphological Geodesic Active Contour

Pedro P. Rebouças Filho, BTS 11
Róger Sarmento, Aldísio
Method for Segmentation of Hemorrhagic Stroke in
Medeiros, Elizângela
BRACIS TECHNICAL
computed tomography image
Rebouças, Lucas Santos SESSION
Predicting the Evolution of COVID-19 Cases and
Tiago Colliri, Alexandre
Deaths Through a Correlations-Based Temporal
Delbem, Liang Zhao
Network

Quantifying Temporal Novelty in Social Networks Victor dos Santos, Rodrigo

using Time-Varying Graphs and Concept Drift Mello, Tatiane Nogueira,
Detection Ricardo Rios

Jackson Souza, Marcelo

Robust Ranking of Brazilian Supreme Court Decisions
Finger

Stocks Clustering Based on Textual Embeddings for André Oliveira, Pedro F. A.

Price Forecasting Pinto, Sérgio Colcher

34
 BTS12 FRIDAY 14h-15h OCTOBER 23

Authorship attribution of Brazilian literary texts Bianca R. Bartolomei,

through machine learning techniques Isabela Drummond

Deep learning models for representing out-of- Johannes Lochter, Renato

vocabulary words Silva, Tiago Almeida

DeepBT and NLP Data Augmentation Techniques: a Taynan Maier Ferreira,

new proposal and a comprehensive study Anna Costa

Dense Captioning using Abstract Meaning Antonio A. Neto, Helena

Representation Caseli, Tiago Almeida BTS 12
Does Twitter data can be used to estimate Reality Edson Matsubara, Lucas BRACIS TECHNICAL
Show outcomes? Rodrigues, Kenzo Sakiyama SESSION
Rafael B. M. Rodrigues,
Entropy-Based Filter Selection in CNNs Applied to Text
Danilo Eler, Wilson
Classification
Marcílio-Jr

Francielle Vargas, Rodolfo

Identifying fine-grained opinion and classifying polarity
Sanches, Pedro Regattiere
on coronavirus pandemic
Rocha

Vinícios Carvalho, Bianca

Machine learning for suicidal ideation identification on
Giacon, Carlos Nascimento,
Twitter for the Portuguese language
Bruno Nogueira

35
 Keynote speaker: Celine Vens
(KULAK, Belgium)
October 23rd 10h

Interpretable models for biological network mining

Networks are omni-present in the biomedical domain: drug-target

interation networks, protein-protein interaction networks and
patient-drug response networks are just a few examples. An
important task in this domain is to predict whether a link exists
Short Bio between two entities. This task can be modelled as a supervised
Celine Vens is an Associate Professor at the machine learning problem. Interpretable models such as decision
faculty of Medicine of KU Leuven, in Bel- gium. trees can lead to novel biological insights by providing an
She obtained her PhD degree in computer
science (machine learning) from the same
explanation for the predictions they make. I will discuss our recent
university. Her research expertise focuses on work on predictive bi-clustering trees, that are specifically designed
multi-output learning (multi-label / multi- to learn from interaction data. As a side product, the decision tree
target / hierarchical prediction), tree based
ensemble learning, survival analysis and provides a complete bi-clustering of the data set. I will further
biological network mining. She has published discuss two extensions that boost the predictive performance:
over 50 research papers in both computer
science and biomedical journals or
constructing an ensemble of predictive bi-clustering trees and
conferences, is member of the editorial board combining them with output space reconstruction methods like
of Machine Learning and DAMI journals, and matrix factorization. Finally, an extension towards multi-label
is programme director for the Biomedical
Sciences programme at KU Leuven campus classification will be described.
Kulak.

36
 Keynote speaker: Yaochu Jin
(University of Surrey, Guildford, U.K.)
October 23rd 11h

Communication Efficient Federated Learning

Short Bio
Yaochu Jin received the B.Sc., M.Sc., and Ph.D. degrees Federated learning is a new distributed learning paradigm
from Zhejiang University, Hangzhou, China, in 1988, 1991,
and 1996, respectively, and the Dr.-Ing. degree from Ruhr
that can preserve data privacy in machine learning. One of
University Bochum, Germany, in 2001. He is currently a
Distinguished Chair, Professor in Computational
the main challenges in federated learning is to reduce the
Intelligence, Department of Computer Science, University
of Surrey, Guildford, U.K., where he heads the Nature
communication costs for transmitting model parameters
Inspired Computing and Engineering Group. His main
research interests include data-driven surrogate-assisted
between the local devices and the central server. This talk
evolutionary optimization, multi-objective evolutionary
learning, trustworthy machine learning, swarm robotics, presents some most recent work on communication
and evolutionary developmental systems.
Dr Jin is presently the Editor-in-Chief of the IEEE
efficient federated learning, including constructing compact
TRANSACTIONS ON COGNITIVE
DEVELOPMENTAL SYSTEMS and the Editor-in-Chief of
AND local models, introducing heterogeneous parameter update,
Complex & Intelligent Systems. He was an IEEE
Distinguished Lecturer and Vice President for Technical
and using ternary quantization. Finally, future directions of
Activities of the IEEE Computational Intelligence Society.
He is the recipient of the 2018 and 2021 “IEEE Transactions
research on federated learning will be briefly discussed.
on Evolutionary Computation Outstanding Paper Award”,
the 2015, 2017, and 2020 “IEEE Computational Intelligence
Magazine Outstanding Paper Award”, and the Best Paper
Award of the 2010 IEEE Symposium on Computational
Intelligence in Bioinformatics and Computational Biology.
He was named by the Web of Science as a “Highly Cited
Researcher in 2019”. He is a Fellow of IEEE.

37
ETS 7
ENIAC TECHNICAL SESSION NATURAL LANGUAGE PROCESSING

ETS 7 FRIDAY 9h-10h OCTOBER 23

Alexandre Inoue, Fábio Cozman,

Automated Emergency Room Triage: Helping Patients Get the Best Treatment
Marcus Prado
Adailton Araujo, Marcos Golo,
From Bag-of-Words to Pre-trained Neural Language Models: Improving Automatic Classification
Breno Viana, Felipe Sanches, Roseli
of App Reviews for Requirements Engineering
A.F. Romero, Ricardo Marcacini
Hate Speech Detection in Portuguese with Naïve Bayes, SVM, MLP and Logistic Regression
Adriano Silva, Norton Roman

Post-processing of machine translation texts based on graph theory Lucas Porto, Evandro Ruiz

Luan Misael, Emanuel Fontelles,

Temporal analysis and visualisation of music
Vinicius Sampaio, Mardônio França

The Winograd Schemas from Hell Fábio Cozman, Hugo Neri

Thiago Ferreira, André L.R. Teixeira,

Towards Fully Automated News Reporting in Brazilian Portuguese João Gabriel M. Campos, Fábio
Cozman, Adriana Pagano
Investigating Sentences Features for Subjectivity and Polarity Classification in Brazilian
Miguel de Oliveira, Tiago de Melo
Portuguese

38
ETS 8
ENIAC TECHNICAL SESSION MULTIAGENT SYSTEMS

ETS 8 FRIDAY 14h-15h OCTOBER 23

Thiago França, Péricles Miranda,

CbDGen: A Complexity-based Synthetic Dataset Generation Tool
Ricardo Prudêncio, André Nascimento
Iam Palatnik de Sousa, Marley M. B. R.
Classificações Explicáveis para Imagens de Células Infectadas por Malária
Vellasco, Eduardo Costa da Silva
Thayanne da Silva, Gustavo Campos,
Classifying Organizational Structures on Targets in the Cooperative Target Observation Raimundo Ferro Jr, Matheus Araújo,
João Andrade, Leonardo F. Costa
Hugo do Nascimento, Rodrigo Filho,
Computational Mining on IBICT BDTD?s Thesis and Dissertation Metadata for Supporting Social
Elismênnia Oliveira, Jordão Nunes,
Science Research
Marcelo Inuzuka
Daniel Ferreira, Diana Adamatti,
Multiagent Simulation to Support Disarmament Policies
Tatiane Bastos
Jean Araujo, Cleyton Rodrigues,
Simulating Indemnity in Civil Suits Through a Description Logic Ontology
Fred Freitas
Ingrid Luana A. Silva, Elaine M. G.
Using the Fuzzy Triangular Naive Bayes to Assess Users in Gynecological Examination Training
Soares, Liliane Machado, Ronei Moraes

An Agent-based Simulation to Study the Spread of COVID-19 in Ibirama (SC) Lucas Teixeira, Fernando Santos

39
 Wagner Meira (mediador)
Professor titular do Departamento de Ciência da Computação da UFMG. Wagner é PhD em
Ciência da Computação pela University of Rochester (1997), além de mestre e bacharel em
Ciência da Computação pela UFMG (1993 e 1990, respectivamente). Atualmente Wagner é
pesquisador em produtividade do CNPq (nível 1B) e sub-coordenador do INCT-Cyber - Instituto
Nacional de Ciência e Tecnologia para uma Sociedade Massivamente Conectada. Publicou mais de

Industry panel 2 três centenas de artigos em periódicos e conferências de impacto e é co-autor dos livros Data
Mining and Analysis - Fundamental Concepts and Algorithms (2014) e Data Mining and Machine
Learning - Fundamental Concepts and Algorithms (2020), publicados pela Cambridge University
October 23rd 15h Press. Seus interesses de pesquisa são em sistemas paralelos e distribuídos, em particular na sua
escalabilidade e eficiência, variando de sistemas massivamente paralelos a plataformas baseadas
na Internet, e em algoritmos de mineração de dados, sua paralelização e aplicação em áreas como
ciência de dados, recuperação de informação, cibersegurança e governança eletrônica.

Sylvio Barbon
Possui graduação em Ciência da Computação pelo Centro Universitário do Norte Paulista
(2005) e Mestrado em Física Computacional pela Universidade de São Paulo (2007),
graduação em Engenharia de Computação pelo Centro Universitário de Votuporanga
(2008) e doutorado (2011) pelo IFSC/USP na área de Física Aplicada Computacional.
Durante 2017 foi professor visitante na Università Degli Studi di Milano e desenvolveu
um projeto de pós-doutorado na Università di Modena e Reggio Emilia. É professor no
curso de Ciência da Computação da Universidade Estadual de Londrina e docente no
Programa de Mestrado em Ciência da Computação da UEL e no Programa de Doutorado
Associado em Engenharia Elétrica. Sua área de pesquisa está concentrada em
Processamento Inteligente de Sinais, Aprendizado de Máquinas, Reconhecimento de
Padrões e Visão Computacional.

Milton Stiilpen Júnior

Mestre em Ciência da Computação, área de atuação Recuperação de Informação e

Análise de Dados Mineração de Dados, atual CTO Stilingue, que é uma plataforma de social listening e
responding desenvolvida para o português Brasil capaz de escutar as conversas online
sobre sua marca e mercado, aplicando inteligência artificial no enriquecimento de dados.

em Tempo de
Pandemia Carlos Renato Lisboa Francês
Professor Titular da Faculdade de Engenharia da Computação e Telecomunicações da UFPA. Pesquisador
Visitante na Chalmers University of Technology (atividades suspensas em função da pandemia Covid-19).
Estágio Pós-Doutoral no Instituto de Engenharia de Sistemas e Computadores, Tecnologia e Ciência
(INESC/Porto) – 2012. Doutor em Ciênca da Computação e Matemática Computacional pelo ICMC/USP –
2001. Orientador dos Programas de Pós-Graduação em Engenharia Elétrica (PPGEE) e de Ciência da
Computação (PPGCC), ambos da UFPA. Possui 21 orientações concluídas de doutorado, 33 de mestrado e
2 supervisões de pós-doutorado. Já publicou 81 artigos em periódicos, 2 livros, 13 capítulos de livro e 176
trabalhos em conferências nacionais e internacionais. Reitor pro-tempore da Universidade Federal do Sul
e Sudeste do Pará (Unifesspa) - 2016. Pró-Reitor de Pós-Graduação, Pesquisa e Inovação Tecnológica
(PROPIT) da Unifesspa - 2013/2016. Presidente da Empresa de Processamento de Dados do Estado do
Pará (PRODEPA) - 2007/2010. Bolsista de Produtividade PQ-1D CNPq.

40
 Organized by

Promotion

Sponsored by

41