We are an interdisciplinary research group that aims to contribute to the understanding of human behavior and its interaction with intelligent systems through cognitive neuroscience, computational modelling, and Brain-Computer Interfaces. The group is developing computational tools in many areas of neuroscience using technologies such as Eye-tracking and EEG. CONCYS researchers come from several backgrounds, such as Computer Science, Linguistics and Psychology.

This interdisciplinary and multi-instituttional project is currently being conducted by the laboratories of Computational Neurosciences and Cybernetics Systems (CONCYS) and of Bilingualism and Cognition ( LABICO ), from the Federal University of Rio Grande do Sul (UFRGS), and the Social Brains Group , from the Vale do Rio dos Sinos University (UNISINOS). the project's main objective is to develop computational tools for solving problems associated with the Portuguese language. Mainly, the research is based on two aspects: the psycholinguistic one, responsible for studying and detecting reading problems through Electroencephalographic (EEG) and Eye-Tracking (ET) in a working memory paradigm; and the Natural Language Processing (NLP) one, which is in charge of creating systems based on NLP for enhancing Human-Computer Interfaces (HCI). Thus, it is expected, on the one hand, to create experimental frameworks and computational models for understanding reading disabilities in Portuguese and, on the other hand, to extend the current NLP systems for Portuguese, mainly question answering (QA) systems.

Question Answering Systems (QA) is a field of Information Retrieval that automatically answers questions posed by humans in a natural language. Currently in the literature, there are QA systems that have achieved good results with machine learning approaches (Sarrouti et al, 2017, Xia et al, 2018) instead of hand-written rules created manually. However, most of them were developed for the English language. Therefore, this research project intends to apply and develop efficient techniques based on machine learning to QA systems for the Portuguese language.

Working Memory (WM) is responsible for briefly maintaining and processing information in order to perform cognitive tasks (decision making, reading, logical operation, etc.); during reading, its job is to temporarily hold words or sentences in memory, match them to the lexicon (vocabulary) and create a logical speech that makes sense for the reader. Savage et. al (2006) demonstrated the straightforward link between WM and attention and reading disorders. However, one of the challenges in working memory is to find out the appropriate paradigm for stimulating the user so that the phenomenon to be studied can be easily identified. Thus, this work aims to develop a system that makes easy the detection of reading disorders based on EEG (ERP-N400) and Eye-Tracking (Fixations -Saccades).

Along of the last decades, the human beings have sought for alternative ways to communicate themselves with machines or systems, in order to overcome difficulties imposed by biological limitations, impairments or just to create a useful wearable. Precisely, Brain-Computer Interface (BCI) employs brain signals (typically, Electroencephalography - EEG) to control or communicate with many systems. This work seeks to explore as much as possible the recently acquired equipment for recording EEG data (OpenBCI) as well as its application in a Virtual Reality (VR) environment. Therefore, a typical BCI methodology is employed for evaluating the device in a motor task using both EEG and Electromyography (EMG) signals in a VR experience where the user can control a virtual arm using either real or imagined movements.

Given recent findings relating low cognitive ability and susceptibility to fake news ( De keersmaecker; Roets, 2017), this research could provide more support to this line of inquiry by exploring the links between low working memory capacity, eye movements, EEG signals and a difficulty in rejecting misinformation.