The article entitled "struc2vec: Learning Node Representations from Structural Identity", co-authored by Leonardo Ribeiro, Pedro Savarese and Daniel Ratton Figueiredo, published in ACM SIGKDD in 2017 reached the milestone of 1000 citations according to Google Scholar.

The research work that led to the article was carried out entirely at the PESC by, at the time, students Leonardo (master's degree) and Pedro (doctorate) and by Professor Daniel. The work was the subject of Leonardo's master's dissertation, defended in June 2017, which won the XI Contest for Theses and Dissertations in Artificial and Computational Intelligence (CTDIAC).

The article deals with the problem of representing vertices of a network in a Euclidean space to capture the local structure of the vertices, regardless of their labels, thus standing out from other proposals in the literature, such as node2vec. The work, which presents an original proposal for the construction of representations and a new evaluation methodology, has publicly available source code (and its own website) and was very well received by the academic community.

Congratulations to the co-authors and to the PESC!