Text Mining, Affective Computing and Behavioral Analysis

This project aims to extract knowledge from different unstructured sources about a given domain. In text mining, information is usually obtained by identifying patterns and trends through statistical or machine learning from texts. In this way, the objective in this area focuses on representing texts in the vector space, developing classification algorithms and building applications aimed at the study domain.

In addition to the meaning of the words and the main message that one wants to convey during the writing, the text produced brings with it different information about the emotions of the one who is writing it. In this context, affective computing is established as computing related to emotion, derived from emotions or that deliberately influences emotions. This area has many open challenges, especially in the area of ​​detecting and classifying emotions. It comprises the study and development of systems that can recognize, interpret, process and simulate human affection. The objective of this project is to develop and use devices capable of recognizing facial expressions, gestures, speech, changes in body temperature, breathing rhythm, among others, as well as extracting significant patterns from these captured data.

Behavioral analysis is the analysis of individuals in groups, whether in social networks, collaboration networks and co-authorship. The objective, in this project, is to analyze the behavior of individuals in their groups or communities modeled by means of graphs with attributes, in order to be able to highlight relevant communities. In this research, different approaches are used, such as grouping algorithms in graphs with attributes for the detection of communities. Among the applications studied are the detection of groups for targeted marketing (target marketing), detection of emotional homophilia (emotion homophily) and the dissemination of information (spread of information).


  • Eduardo Bezerra
  • Kele Belloze
  • Gustavo Guedes (Coordinator)

International partnerships:

  • Mark Turner (Case Western Reserve University)
  • Francis Steen (UCLA)

Financial Information:

  1. FAPERJ APQ1 announcement, project “Multiple non-redundant groupings in graphs with attributes”, in the 2016-Current period, with the coordination of Professor Gustavo Guedes. Financed amount: R$ 8,500.00;
  2. FAPERJ Notice Installation, in the 2017-2018 period, with the coordination of Professor Gustavo Guedes. Financed amount: R$ 6,950.00;
  3. PIBIC and PIBIC-EM scholarships.

These projects are being developed by the group members and total a financing amount of approximately R$ 15,450.00.

Comments are closed.