Utilização de redes para a deteção de casos de fraude em apólices de seguro automóvel

O tema da fraude nas seguradoras tem sido um objecto de estudo, durante as recentes décadas, por investigadores de diferentes áreas da sociedade, tanto do meio académico como por especialistas da indústria, reflectindo assim a importância deste tema no campo da economia e da aceitação social. O principal objectivo teórico é a criação de uma rede que revele as relações entre uma base de dados de dados de sinistros segmentada, sendo esta providenciada por três grandes seguradoras que operam no mercado dos seguros não-vida português, obtendo assim informação sobre as características do cliente fraudulento. Mais especificamente, os segmentos da análise são feitos através da análise de duas variáveis alvo presentes nesta que dão informação se o cliente é não-suspeito e não-confirmado, não-suspeito e confirmado, suspeito e não confirmado e suspeito e confirmado. Após a análise, os quatro segmentos são criados e posteriormente, a rede é elaborada com base na relação das diferentes características dos segmentos criados. Os resultados mostram que os segmentos que se referem aos casos suspeitos e não confirmados e os casos suspeitos e confirmados são os que mais se relacionam, por outro lado, o segmento dos casos não suspeitos e não confirmados é o que menos características em comum.

As 'conversas informais' extraprocessuais como meio de prova em processo penal

In the present thesis, we examine the approach to the so-called “informal conversations”, especially between a suspect or defendant and criminal police authorities. Our goal is to understand if criminal police authorities are allowed to testify about the content of these conversations, revealing facts that the suspect or defendant may have shared with them, as well as about evidence that they may have acquired through these statements. Firstly, we briefly present the notion of “informal conversations” and the great variety of situations they may encompass: intra or extra-procedural; prior or subsequent to someone acquires the status of defendant. Secondly, we analyse some of the principles and rules that are involved in this controversial issue: principles concerning the procedural structure, organization and dynamic; principles concerning the production and assessment of evidence in the trial hearing; principles concerning the prosecution and the powers of criminal police authorities; the procedural status of the defendant; the rules concerning the reading of statements in the trial hearing; the rules concerning hearsay testimonies. Thirdly, we go through the great amount of case law on the so-called “informal conversations” and related matters, analysing the most relevant cases and the arguments that sustain them, as well as the legal literature. Our goal is to understand the evolution, throughout the last two decades, of the different opinions regarding the approach to the various situations in which “informal conversations” may occur and in which the admissibility of a testimony by criminal police authorities is questioned. Finally, we defend a different approach for testimonies by criminal police authorities prior and subsequent to someone acquiring the status of defendant. We see the moment when someone acquires the status of defendant as a border area in the admissibility of “informal conversations”, because from then on the statements have to be collected and assessed according to the law, so all the other conversations (or any other evidence) collected informally are irrelevant. As to the specific case of the testimony about the re-enactment of the crime, given the high degree of difficulty in separating the defendant’s contributions that may be considered essential and those that may be considered less useful, but still relevant, we support the qualification of the defendant’s contributions as inseparable from the re-enactment, allowing it to be replicated and assessed in the trial hearing with no restrictions.