Banca de QUALIFICAÇÃO: WELLINGTON BARBOSA RODRIGUES

Uma banca de QUALIFICAÇÃO de DOUTORADO foi cadastrada pelo programa.
STUDENT : WELLINGTON BARBOSA RODRIGUES
DATE: 19/09/2022
TIME: 15:00
LOCAL: Remoto, via GoogleMeet https://meet.google.com/ckq-uymh-kbw
TITLE:

Computational methods for categorizing conflicts of interest in peer review

PAGES: 40
BIG AREA: Ciências Exatas e da Terra
AREA: Ciência da Computação
SUBÁREA: Metodologia e Técnicas da Computação
SUMMARY:

The identification of reviewers for peer review is a recurring task in the science construction process and is a big challenge for the people responsible for organizing the evaluation boards. To define reviewers, the criteria determined by the process in which the peer review will be applied must be followed, and there may be variations between different processes, however, some criteria are common to practically all evaluations, for example, the exemption of conflicts of interest. When referring to the problem of conflict of interest, it is difficult to be subjective, currently the best accepted means of identifying conflicts of interest in peer review is self-declaration. However, one of the difficulties of self-declaration is that it will only be possible to use it after the invitation of possible reviewers. Another relevant point is that the first proposal of a peer review of organizer tends to be for the closest and known people, with a propensity to generate an accumulation of invitations to some reviewers and fewer invitations than desired for others. Most peer reviews are done by a small group of people.

This work proposes the creation of methods for categorizing conflicts of interest, which makes it possible to identify possible reviewers for peer review. The objective is to analyze the history of previous collaborations and define an algorithm that can categorize potential conflicts of interest and present the most suitable researchers to carry out a review. Thus, streamlining the selection process and enabling a better pluralization of possible reviewers.
The proposal consists of creating a graph of collaboration of researchers and segmenting it into smaller subgraphs, according to the context that will be applied, thus identifying the intensity of collaboration in the context. After this segmentation, suggest a group of researchers less subject to potential conflicts of interest. As preliminary results, experiments were carried out to categorize the relationships of researchers with a CNPq research productivity grant. The initial experiments are divided by area of knowledge and by periods. In this way, it is possible to verify the nuances of each area of knowledge in this group of researchers. It was identified that there is a large discrepancy between the patterns of collaboration and, therefore, requiring a dynamic and contextualized analysis to identify potential conflicts of interest in each peer review process.

BANKING MEMBERS:
Presidente - Interno ao Programa - 1934625 - JESUS PASCUAL MENA CHALCO
Membro Titular - Examinador(a) Externo à Instituição - SAMILE ANDRÉA DE SOUZA VANZ - UFRGS
Membro Titular - Examinador(a) Externo à Instituição - LUCIANO ANTONIO DIGIAMPIETRI - USP
Membro Suplente - Examinador(a) Externo à Instituição - FABIO CASTRO GOUVEIA - FIOCRUZ - RJ