Markov Constraints for Generating Texts with Style

This thesis addresses the issue of generating texts in the style of an existing author, that also satisfy structural constraints imposed by the genre of the text. Although Markov processes are known to be suitable for representing style, they are difficult to control in order to satisfy non-local properties, such as structural constraints, that require long distance modeling. The framework of Constrained Markov Processes allows to precisely generate texts that are consistent with a corpus, while being controllable in terms of rhymes and meter. Controlled Markov processes consist in reformulating Markov processes in the context of constraint satisfaction. The thesis describes how to represent stylistic and structural properties in terms of constraints in this framework and how this approach can be used for the generation of lyrics in the style of 60 differents authors An evaluation of the desctibed method is provided by comparing it to both pure Markov and pure constraint-based approaches. Finally the thesis describes the implementation of an augmented text editor, called Perec. Perec is intended to improve creativity, by helping the user to write lyrics and poetry, exploiting the techniques presented so far.

Hybrid Artificial Intelligence to Extract Patterns and Rules from Argumentative and Legal Texts

This Thesis is composed of a collection of works written in the period 2019-2022, whose aim is to find methodologies of Artificial Intelligence (AI) and Machine Learning to detect and classify patterns and rules in argumentative and legal texts. We define our approach “hybrid”, since we aimed at designing hybrid combinations of symbolic and sub-symbolic AI, involving both “top-down” structured knowledge and “bottom-up” data-driven knowledge. A first group of works is dedicated to the classification of argumentative patterns. Following the Waltonian model of argument and the related theory of Argumentation Schemes, these works focused on the detection of argumentative support and opposition, showing that argumentative evidences can be classified at fine-grained levels without resorting to highly engineered features. To show this, our methods involved not only traditional approaches such as TFIDF, but also some novel methods based on Tree Kernel algorithms. After the encouraging results of this first phase, we explored the use of a some emerging methodologies promoted by actors like Google, which have deeply changed NLP since 2018-19 — i.e., Transfer Learning and language models. These new methodologies markedly improved our previous results, providing us with best-performing NLP tools. Using Transfer Learning, we also performed a Sequence Labelling task to recognize the exact span of argumentative components (i.e., claims and premises), thus connecting portions of natural language to portions of arguments (i.e., to the logical-inferential dimension). The last part of our work was finally dedicated to the employment of Transfer Learning methods for the detection of rules and deontic modalities. In this case, we explored a hybrid approach which combines structured knowledge coming from two LegalXML formats (i.e., Akoma Ntoso and LegalRuleML) with sub-symbolic knowledge coming from pre-trained (and then fine-tuned) neural architectures.

Women Writers and Women’s Rights: The Denunciation of Women’s Conditions in the Texts of Mary Leman Grimstone and Gertrudis Gómez de Avellaneda

Through the analysis of the prose of two nineteenth-century women writers: the English Mary Leman Grimstone and the Cuban-Spanish Gertrudis Gómez de Avellaneda, the present dissertation aims at unveiling the relationship between women’s writings and the struggle for the recognition of women’s rights in two different geopolitical locations. To do so, it weaves a Feminist Planetary Web between each writer and her context, as well as among both writers, finding points of connection and disconnection. It shows how women appropriated the pen in different geographical locations, exposing a particular female voice that denounced not only the oppression suffered by women, but also by other marginalized subjects. For each writer this dissertation exposes several macro-arguments present transversally in their work, like their critiques to the institution of marriage, the importance of proper education for women, the advocacy for religious tolerance, and the narrative construction of different male and female paradigms. The critiques to the institution of marriage is a point of connection between both authors. They also coincided in highlighting the importance of women’s right to access a proper education. Aside from these commonalities, this dissertation also analyses how Grimstone and Gómez de Avellaneda negotiated their position in the literary public realm, showing how it was precisely in this point that readers and critics can find noteworthy differences between them.