On December 15, 2023 at 11:00 EET (FMI, Hall 214 “Google”), we will have two short talks (30 min. each) in the Data Science Seminar, by Miruna Zăvelcă and Eduard Szmeteanca:
Miruna Zăvelcă (PhD student at the University of Bucharest)
Establishing Language-Agnostic Causality Trees on Argumentative Texts
Abstract:
In recent years, argumentation mining has been receiving increasing attention in the research world. With recent research trends focusing on the correlation between causality and emotions, it is important to understand the background on the grammatical particularities of arguments, argumentative essays and stance detection, especially considering the strong link between sentiment, opinion, and argumentative structure (Hogenboom et al., 2010). In this seminar I explain these concepts, then use them to create a prototype for a data processing pipeline that takes an essay as input and identifies and builds a causality tree based on the argumentative components in it. Everything is designed to be language-agnostic to help fill the gap that the field of Natural Language Processing has in non-English languages, including Romanian.
and
Eduard Szmeteanca (PhD student at the University of Bucharest)
Raman spectroscopy for tumor detection
Abstract:
Thanks to technological progress, Raman spectroscopy is increasingly used in biological studies, and more than that, it can be used to detect tumors, thus shortening the time required for diagnosis and at the same time increasing the possibilities of treatment. In addition, with the help of Raman spectroscopy, the boundaries of a tumor can be identified so that, during surgery, decisions can be made about its removal. This is especially beneficial in the case of tumors where operating a larger portion than necessary may mean restricting the patient’s capabilities as happens for example in brain cancer. This seminar will provide an introduction to research in this field.