no code implementations • 7 Jan 2024 • Julia El Zini, Bassel Musharrafieh, Mariette Awad
In this work, we investigate the features that are learned by deep models and we study whether these deep networks are able to encode features as complex and high-level as the fractal dimensions.
1 code implementation • 19 Jan 2023 • Julia El Zini, Mohammad Mansour, Mariette Awad
Our Contrastive Entropy-based explanation method, CEnt, approximates a model locally by a decision tree to compute entropy information of different feature splits.
1 code implementation • 17 Oct 2022 • Julia El Zini, Mariette Awad
Accordingly, we extend the computation of three metrics, proximity, connectedness and stability, to textual data and we benchmark two successful contrastive methods, POLYJUICE and MiCE, on our suggested metrics.
no code implementations • 13 Oct 2022 • Julia El Zini, Mariette Awad
Such challenges can be attributed to the lack of input structure in textual data, the use of word embeddings that add to the opacity of the models and the difficulty of the visualization of the inner workings of deep models when they are trained on textual data.
no code implementations • 13 Oct 2022 • Julia El Zini, Mohamad Mansour, Basel Mousi, Mariette Awad
In this work, inspired by offline information retrieval, we propose different metrics and techniques to evaluate the explainability of SA models from two angles.
no code implementations • 26 Nov 2019 • Julia El Zini, Yara Rizk, Mariette Awad
Recurrent neural networks (RNN) have been successfully applied to various sequential decision-making tasks, natural language processing applications, and time-series predictions.
no code implementations • 7 Nov 2019 • Rema Daher, Mohammad Kassem Zein, Julia El Zini, Mariette Awad, Daniel Asmar
Transfer learning improves the GV by 35% and the MS by 13% on average.
no code implementations • COLING 2016 • R El Khatib, a, Julia El Zini, David Wrisley, Mohamad Jaber, Shady Elbassuoni
TopoText calculates the number of times a place was mentioned in the text, which is then reflected on the map allowing the end-user to grasp the importance of the different places within the text.