Lab CORES is an initiative of the Universidade Federal do Rio de Janeiro.
Address
Av. Athos da Silveira Ramos, 274
Ilha da Cidade Universitária
Rio de Janeiro - RJ
Brasil
Follow us on our social networks
Domains
Urban Centers
Scenario that characterizes applications and solutions aimed at solving problems and creating facilities in urban centers.
Research Topics: Research Topics: Smart Cities; Crowdsensing; Opportunistic collaborations(opportunistic sensing, opportunistic networks); Urban mobility.
Science / Academy
Scenario featuring applications and solutions geared to academia and scientific research. It mainly includes researches focused on analyzing scientific production, technologies and products, as well as the interaction between scenario and society.
Research Topics: Research Topics: Scientometrics (metrics on science and production); Altmetrics (Alternative metrics that aim to quantify other aspects of science, eg impact on the population); Tech mining; Big scholar data (Techniques for application in massive academic data).
Areas
Intelligent Ecosystems
Research area dedicated to the development of models, techniques and computational practices focused on intelligent ecosystems. An ecosystem is a metaphor used to represent collaborative arrangements in constant evolution. These communities have autonomous entities with distinct purposes, creating a complex network of relationships for coexistence.
Through social computing resources, different kinds of intelligent ecosystems are analyzed, such as entrepreneurial communities, software development, innovation organizations, and smart cities. Among other purposes, the researches aim to systematize the decision-making process and improve the mechanisms of knowledge management, transparency, and resilience of these environments.
Research Topics: Education; Entrepreneurship and Innovation; Knowledge Management; Decision Support Systems; Emergency; Governance; Team Formation; Fairness, Accountability, Transparency, and Ethics on the Web.
Sensing and Monitoring
Through social media, people share opinions, experiences and perspectives, which can be revealed by different formats such as text, images, audio and video. Also, we can identify unprecedented levels of citizen engagement and participation. Users can detect new events, help in dealing with natural disasters, diseases, terrorism, and understand the political scenario and support public manifestations. Citizens consume and share information about their cities - such as problems, events, ideas, suggestions, criticisms, and demands – acting as ‘human sensors’, forming opinions and participating in the city evolution. This activity is called “participatory sensing” or “citizen sensing”, a form of citizen participation in environmental monitoring and action which is bottom-up, empowering a community.
Moreover, sensors are embedded into our everyday environment, such as cars, micro-chipped animals, elevators, smart buildings, environmental monitors and others. Moreover, many people are now carrying some form of sensor-laden device - a mobile phone, a tablet, a fitness device - from which sensor readings can also be retrieved. In this case, people interact – direct and indirect – with sensors. This is sometimes called ‘human-in-the-loop sensing’, empowering the citizen sensing with new kind of data. In citizen sensing, a sensor is not necessarily a hardware sensor, but can be a virtual sensor or a human interpreting sensory data.
Research Topics: Social IoT; Mobile and New types of human interactions; Event identification; Problems in urban centers (mobility, diseases, etc); Emotion and Sentiment Analysis; Profiling; Hate speech in social media; Fake news, social bots, misinformation, and disinformation on social media; Credibility and reputation in social media, Media Literacy.
Pattern Recognition
This group is responsible for the creation and application of statistical and computational methods for pattern recognition on structured and unstructured data like texts, images and time series. This module is responsible for providing solutions to the other modules for the creation of intelligent applications.
Research Topics: Pattern recognition in streaming data; Learning from few examples (one-shot learning); Learning from unbalanced datasets; Model interpretability; Semantic extraction from unstructured data (images, time series, audio, etc.); Data treatment for bias removal; Topic discovery (events, subjects) in semi-structured and unstructured data; Information Retrieval; Recommendation.
Data Engineering and Semantic Treatment
In Big Social Data context, the massive growth of data demands new techniques to access, manage, mine, analyze, and act on data. Real-time mining of indirectly self-reported and sousveillance information harvested from social media can offer useful data and insights about unfolding trends and emerging crowd behaviors, especially at times of crises. However, such data often contain variable amounts of “noise”, misinformation and bias (which can get further “amplified” through the viral nature of social media) and usually require some advanced forms of filtering and verification by both machine-based algorithms and human experts before becoming reliable enough for use in decision-making tasks. Eradicating these issues or lessening their effect can be accomplished by verification of data with other sensor nodes, but depends on the density of the network and existence of other related data.
So, this module is responsible for the development of techniques to manage the massive volume of social data, envisioning the urban sensing, dealing with multiple data streams of varying data types, and different knowledge levels of users.
Research Topics: Benchmarking, Performance Modelling, and Tuning; Data Integration, Metadata Management, and Interoperability; Data Models, Semantics, Query languages; Social Data Provenance, Cleaning, and Curation; Data Stream Systems and Sensor Networks; Data Visualization and Interactive Data Exploration; Privacy, Security, and Trust; Search and Information Extraction; Temporal, Spatial, Mobile, and Multimedia.
Computational Ethics
We focus on ethics applied to the digital universe with a specific lens aimed at, but not limited to, the Brazilian context. Our investigations also have a meta-scientific bias, covering in parallel the collaborative networks between researchers and how they form, evolve and influence scientific production. We study the ethical or moral aspects and implications that emerge on digital platforms, including academic interactions. Recognizing the richness and diversity of the Brazilian scenario, we seek to ensure that research and collaboration practices in the digital environment are conducted with integrity, transparency and respect. We are dedicated to building a solid knowledge base that serves as a guide for ethical conduct in the Brazilian digital landscape, valuing and promoting best practices in our academic and research environment.
Research Topics: Applied Ethics; Computational Ethics; Research Ethics (all of which are dedicated to the analysis and mining of social networks or social computing); Meta-Science; Critical Theory; Transparency
Toxic Digital Content
This area aims to create solutions to identify and combat the spread of harmful content on social media, such as disinformation, hate speech, misogyny, transphobia and pedophilia. Identification is done by analyzing and detecting patterns of content characteristics (syntax, semantics, subjects covered, layout, as well as feelings and emotions expressed) and context (profile characteristics, propagation, interactions about the content, etc.). Once toxic content has been identified, one of the ways to combat it is to understand the path of this information. Multi-network propagation refers to the spread of information (in this case, toxic information) through various interconnected networks such as online social networks, messaging tools, news networks, forums and other communication channels. Modeling propagation in this context can help us understand how this information spreads, how it can become viral and what factors influence its spread. We know that combating this type of content is not only done through automatic or semi-automatic techniques. That is why we have also created mechanisms to support media literacy, teaching people how to identify the characteristics of toxic content and how to combat it.
Research Topics: Multi-network Information Propagation Models; Network Topologies; Influencer Identification; Textual and Contextual Pattern Identification; Community Identification; Team Formation; Information Provenance; Media Literacy; Crowdsourcing.