Automatic extraction of concepts from texts and applications

The extraction of relevant terms from texts is an extensively researched task in Text- Mining. Relevant terms have been applied in areas such as Information Retrieval or document clustering and classification. However, relevance has a rather fuzzy nature since the classification of some terms as relevant or not relevant is not consensual. For instance, while words such as "president" and "republic" are generally considered relevant by human evaluators, and words like "the" and "or" are not, terms such as "read" and "finish" gather no consensus about their semantic and informativeness. Concepts, on the other hand, have a less fuzzy nature. Therefore, instead of deciding on the relevance of a term during the extraction phase, as most extractors do, I propose to first extract, from texts, what I have called generic concepts (all concepts) and postpone the decision about relevance for downstream applications, accordingly to their needs. For instance, a keyword extractor may assume that the most relevant keywords are the most frequent concepts on the documents. Moreover, most statistical extractors are incapable of extracting single-word and multi-word expressions using the same methodology. These factors led to the development of the ConceptExtractor, a statistical and language-independent methodology which is explained in Part I of this thesis. In Part II, I will show that the automatic extraction of concepts has great applicability. For instance, for the extraction of keywords from documents, using the Tf-Idf metric only on concepts yields better results than using Tf-Idf without concepts, specially for multi-words. In addition, since concepts can be semantically related to other concepts, this allows us to build implicit document descriptors. These applications led to published work. Finally, I will present some work that, although not published yet, is briefly discussed in this document.

A intervenção dos orgãos de polícia criminal em sede de julgamento

Currently, Portugal assumes itself as a democratic rule of substantive law State, sustained by a legal system seeking the right balance between the guarantee of fundamental rights and freedoms constitutional foreseen in Portugal’s Fundamental Law and criminal persecution. The architecture of the penal code lies with, roughly speaking, a accusatory basic structure, “deliberately attached to one of the most remarkable achievements of the civilizational democratic progress, and by obedience to the constitutional commandment”, in balance with the official investigation principle, valid both for the purpose of prosecution and trial. Regarding the principle of non self-incrimination - nemo tenetur se ipsum accusare, briefly defined as the defendant’s right of not being obliged to contribute to the self-incrimination, it should be stressed that there isn’t an explicit consecration in the Portuguese Constitution, being commonly accepted in an implicit constitutional prediction and deriving from other constitutional rights and principles, first and foremost, the meaning and scope of the concept of democratic rule of Law State, embedded in the Fundamental Law, and in the guidelines of the constitutional principles of human person dignity, freedom of action and the presumption of innocence. In any case, about the (in) applicability of the principle of the prohibition of self-incrimination to the Criminal Police Bodies in the trial hearing in Court, and sharing an idea of Guedes Valente, the truth is that the exercise of criminal action must tread a transparent path and non-compliant with methods to obtain evidence that violate the law, the public order or in violation of democratic principles and loyalty (Guedes Valente, 2013, p. 484). Within the framework of the penal process relating to the trial, which is assumed as the true phase of the process, the witness represents a relevant figure for the administration of criminal justice, for the testimonial proof is, in the idea of Othmar Jauernig, the worst proof of evidence, but also being the most frequent (Jauernig, 1998, p. 289). As coadjutant of the Public Prosecutor and, in specific cases, the investigating judge, the Criminal Police Bodies are invested with high responsibility, being "the arms and eyes of Judicial Authorities in pursuing the criminal investigation..." which has as ultimate goal the fulfillment of the Law pursuing the defense of society" (Guedes Valente, 2013, p. 485). It is in this context and as a witness that, throughout operational career, the Criminal Police Bodies are required to be at the trial hearing and clarify the Court with its view about the facts relating to occurrences of criminal context, thus contributing very significantly and, in some cases, decisively for the proper administration of the portuguese criminal justice. With regards to the intervention of Criminal Police Bodies in the trial hearing in Court, it’s important that they pay attention to a set of standards concerning the preparation of the testimony, the very provision of the testimony and, also, to its conclusion. Be emphasized that these guidelines may become crucial for the quality of the police testimony at the trial hearing, thus leading to an improvement of the enforcement of justice system. In this vein, while preparing the testimony, the Criminal Police Bodies must present itself in court with proper clothing, to read before and carefully the case files, to debate the facts being judged with other Criminal Police Bodies and prepare potential questions. Later, while giving his testimony during the trial, the Criminal Police Bodies must, summing up, to take the oath in a convincing manner, to feel comfortable, to start well by convincingly answering the first question, keep an attitude of serenity, to adopt an attitude of collaboration, to avoid the reading of documents, to demonstrate deference and seriousness before the judicial operators, to use simple and objective language, to adopt a fluent speech, to use nonverbal language correctly, to avoid spontaneity responding only to what is asked, to report only the truth, to avoid hesitations and contradictions, to be impartial and to maintain eye contact with the judge. Finally, at the conclusion of the testimony, the Criminal Police Bodies should rise in a smooth manner, avoiding to show relief, resentment or satisfaction, leaving a credible and professional image and, without much formality, requesting the judge permission to leave the courtroom. As final note, it’s important to stress that "The intervention of the Police Criminal Bodies in the trial hearing in Court” encloses itself on a theme of crucial importance not only for members of the Police and Security Forces, who must welcome this subject with the utmost seriousness and professionalism, but also for the proper administration of the criminal justice system in Portugal.

Social tariffs for water and waste services in mainland Portugal: An impact analysis

This paper proposes to quantify the effect of social tariffs (ST) in the Portuguese water and waste sector (WWS). It calculates the amount of subsidy implicit in ST schemes, characterising the existing tariffs in 2011 and producing a synthetic tariff scene where the regulator’s recommendation is respected. This is the first time such an exercise is undertaken and it is very relevant in a context of deep economic crisis. Results suggest that there are fewer beneficiaries than what income eligibility criteria would imply and that putting the regulator’s recommendation in practice would considerably raise subsidy amounts, potentially leading to a severe increase in non-subsidised user tariffs to allow for break-even.

Aspects of Passives in Capeverdean (Santiago’s variety)

The present dissertation intends to study passives in Capeverdean. I argue that Capeverdean have eventive passives with specific morphology, postverbal morphemes, -du and -da, which affix to the verb to form passives and interact with the TMA morphemes available in the language in the same way as in the active voice. I also show that Capeverdean only allows short passives. However, this study demonstrates that the by-phrase, although not expressed phonologically, is implicit and can be tracked through instrumentals and Agent-oriented adverbs. In order to account for this specific property of Capeverdean passives, I assume the existence of a Voice head which introduces the external argument in all finite sentences in Capeverdean, except in unaccusatives, following proposals from Marantz (1984), Kratzer (1996), Sailor & Ahn (2010), Pratas (2014). I also assume that this Voice head is subject to a Doubly Filled Comp Filter, similar to what is proposed in Koopman (1997), which determines that either heads or specifiers can be overt, never both. In the case of passives, I propose that external argument is in Spec,Voice and the passive morphology is lexicalized in Voice0 and that while Spec,Voice is silent, Voice0 is not. This configuration can be explained if it is assumed, following Costa & Martins (2004), that in Capeverdean passives Voice0 is a strong functional head, thus requiring visibility at PF. This restriction, combined with the Doubly Filled Comp filter, imposes that Spec,Voice is silent.

Lovers and Madmen: The Μανία-Φρονεῖν Opposition in Plato’s Phaedrus

In this thesis we propose to examine the first half of the Phaedrus (sc. until the end of the palinode) in light of the opposition between the notions of μανία and φρονεῖν, as they are explicitly and implicitly presented in the erotic speeches. These are read in dialogue with what we have designated as the “implicit speech” or “speeches”, i.e., the plurality of conceptions regarding ἔρως, μανία and φρονεῖν that were part of Ancient Greek culture. Our reading of the two speeches against ἔρως, Lysias’ and Socrates’ first speech, engages with this cultural background, and extracts a conception of μανία and φρονεῖν with which the palinode will primarily confront. Our reading of the palinode divides it into two sections: the first, the presentation of the first three kinds of beneficial μανία; the second, the mythical narrative that deals with erotic μανία. We emphasise the existence of a wide gulf between these two moments in terms of their ontological, theological and anthropological conceptions. The second section of the palinode is revolutionary not only in contrast with the “implicit speech” and the speeches against ἔρως, but also in contrast with the very beginning of the palinode – which preserves many of the conceptions and assumptions found in the previous speeches and in the cultural tradition. It is in order to explain the foundation, meaning and significance of this gulf that we explore and discuss the notion of ὑπόθεσις and its role as an implicit operator in the Phaedrus. From our reading of the second part of the palinode, it is clear how the introduction of the ὑπερουράνιος τόπος brings about a radical revision of the perspectives on the nature of reality and on human nature and condition that were implicit in the previous speeches and in the first part of the palinode. We show that the ὑπερουράνιος τόπος corresponds to the projection of a multiplicity of cognitive and desiderative requirements that our normal perspective demands, but cannot possibly satisfy. In other words, our perspective is shown to be living beyond its means, yearning for something that by far exceeds what it can get in its de facto condition: the superlative. This results in a major revision of the understanding of φρονεῖν and μανία – a revision that challenges the traditional understanding of these two notions as binary opposites, thereby revealing a much more complex landscape.

Visual politics of the estado novo realized on the Alameda Dom Afonso Henriques in Lisbon – the construction of ideal citizenship?

The stylistic categorization of the Estado Novo has been intensely discussed by Portuguese art historians. The square Alameda Dom Afonso Henriques in Lisbon (Alameda) can be seen as paradigmatic for the architecture of power of the Estado Novo. The Alameda forms a gardened valley between two hills. There you find two prominent and highly propagandist buildings: The Instituto Superior Técnico (IST) and the Fonte Luminosa are dedicated to modern sciences and respectively to the harmonious contribution of nature to the city. The iconography of the Alameda as well as its incorporation into the propagandist use of urban planning in the 1930s and 1940s exemplify the visual politics during Salazarism. Urban planning programs intended to create cities that would preserve the character of a traditional catholic society and at the same time answer to the need to modernize the country and evoke the image of a progressive state. Thus, public buildings and urban squares such as the Alameda contributed to design a corporate image and to the ‘spirit’ of the regime.

Avaliação do risco de LMEMSLT : aplicação dos métodos RULA e SI / Florentino , António de Sousa Uva

RESUMO - A frequente insuficiência de informação e até, eventualmente, a presença de elementos cientificamente pouco robustos (ou mesmo hipotéticos) no procedimento de obtenção dos resultados finais (scores) com os diversos métodos de avaliação do risco, julga-se determinar a obtenção de distintos resultados no mesmo posto de trabalho e que interessam analisar. O presente estudo foi efectuado numa empresa da indústria automóvel e pretendeu contribuir para a efectividade do processo de avaliação do risco destas lesões em meio industrial. Utilizaram-se como ponto de partida as classificações de risco em postos de trabalho onde a aplicação do método OCRA — base do projecto de norma Europeu prEN 1005-5 para a estimativa do risco de LMEMSLT (Lesões Musculoesqueléticas do Membro Superior Ligadas ao Trabalho) — identificou níveis de risco moderados e/ou elevados (score OCRA ≥ 16,5). Assim nos postos de trabalho (n = 71) registou-se em vídeo a actividade de trabalho. Analisaram-se os principais factores de risco (postura, força, repetitividade e exposição a vibrações) com uma amostragem de segundo em segundo e aplicaram-se dois métodos: 1) RULA (McAtamney; Corlett, 1993); 2) SI (Moore; Garg, 1995). Dos resultados globais da aplicação dos métodos identificam- se discrepâncias evidentes: o método SI classificou 41 postos de risco elevado e o método RULA classificou apenas 26 postos de risco. Destaque para as divergências entre os postos classificados de risco elevado como, por exemplo: de entre os 41 postos com scores de risco da aplicação do método SI apenas se encontram 12 postos classificados de risco pelo método RULA. Os resultados permitem evidenciar validades preditivas diversas para os factores de risco analisados: 1) situações de aplicação de força (SI = 0,80; RULA = 0,66); 2) presença de posturas extremas (SI = 0,68; RULA = 0,48); 3) repetitividade (SI = 0,35; RULA = 0,43). Conclui-se pela divergência de resultados dos métodos SI e RULA aplicados nos mesmos postos de trabalho que tal situação alerta para a pertinência da utilização de um filtro que permita a identificação dos factores de risco presentes em cada posto de trabalho e que, por consequência, oriente a selecção do método mais indicado ou ainda, em oposição, do método contra-indicado. Este trabalho, como um contributo para uma efectiva identificação e avaliação do risco de LMEMSLT, conduz, no limite, à necessidade de elaboração de uma «grelha» das «grelhas» e de informação mais precisa sobre os métodos e sua aplicação, permitindo, desse modo, uma mais efectiva gestão do risco destas lesões.

Uncertainty in olfactory decision-making

Relationships between accuracy and speed of decision-making, or speed-accuracy tradeoffs (SAT), have been extensively studied. However, the range of SAT observed varies widely across studies for reasons that are unclear. Several explanations have been proposed, including motivation or incentive for speed vs. accuracy, species and modality but none of these hypotheses has been directly tested. An alternative explanation is that the different degrees of SAT are related to the nature of the task being performed. Here, we addressed this problem by comparing SAT in two odor-guided decision tasks that were identical except for the nature of the task uncertainty: an odor mixture categorization task, where the distinguishing information is reduced by making the stimuli more similar to each other; and an odor identification task in which the information is reduced by lowering the intensity over a range of three log steps. (...)