Speaker: Vanessa Queiroz Marinho
University of Toronto and University of São Paulo, São Carlos
: Complex networks for text analysis
Complex networks have proven useful to model several real systems of very distinct nature. The discovery that methods from complex networks can be used to analyse texts has allowed the study of NLP tasks from a new perspective. Examples of tasks studied via topological analysis of networks are keyword identification, automatic extractive summarization and authorship attribution. In this talk, I’ll present my work on authorship attribution using methods of complex networks. In addition, we have introduced an approach that combines both traditional techniques with properties provided by the topological analysis of complex networks. In this last approach, we applied the concept of motifs, recurrent interconnection patterns, in the authorship attribution task. The results showed that motifs are able to distinguish the writing style of different authors. Even though such approaches have not outperformed other traditional models, the results obtained with them are quite high, showing that the style of different authors leave fingerprints on the networks derived from their texts.
Special feature: Brazilian desserts by Vanessa