Induction of Topic-Based Bayesian Networks from Text for the Prediction of Sugar C...
Conformed Thought and Art in the Context of Algorithmic Logic Procedures and AI.
Logprob: probabilistic logic --- foundations and computational applications
Grant number: | 19/26762-0 |
Support Opportunities: | Scholarships in Brazil - Post-Doctoral |
Start date: | March 01, 2023 |
End date: | September 30, 2024 |
Field of knowledge: | Physical Sciences and Mathematics - Computer Science - Computing Methodologies and Techniques |
Principal Investigator: | Fabio Gagliardi Cozman |
Grantee: | Paulo Pirozelli Almeida Silva |
Host Institution: | Escola Politécnica (EP). Universidade de São Paulo (USP). São Paulo , SP, Brazil |
Abstract The objective of this research is to develop a methodology that allows the detection of typical patterns of argumentation in texts, such as modus ponens, modus tollens and syllogisms. In order to do this, I will combine resources from logic and argumentation theory with natural processing language techniques and neural network models. The methodology will be based on supervised learning, with training in two corpora created especially for this task: one with examples taken from manuals of logic and argumentation theory; and another, made up of newspaper editorials. The documents in those corpora will have markings for the typical argumentative structures mentioned, formalized in both propositional calculus and first-order logic. Next, I will also add information of syntactic nature (the lexical categories) and semantics (embeddings). The idea is to develop a model that allows these patterns of argumentation to be identified automatically, and which can be further employed in the analysis of the form and content of arguments of natural language, such as in reviews of scientific literature, political debates or court decisions. | |
News published in Agência FAPESP Newsletter about the scholarship: | |
More itemsLess items | |
TITULO | |
Articles published in other media outlets ( ): | |
More itemsLess items | |
VEICULO: TITULO (DATA) | |
VEICULO: TITULO (DATA) | |