Iskanje po katalogu digitalne knjižnice

Opis: The master’s thesis is devoted to the issue of Artificial Intelligence (AI) perspectives in the judiciary, in particular, its application to selected parts of civil proceedings. AI affects virtually the future of every industry and every human being. The application of AI technologies in the legal industry is an issue of growing interest. In particular, attention is drawn to the judicial system due to the fact that, apart from its position of guarantor of justice in society enabling its members to enjoy their rights and freedoms granted by law, it is a service of its nature. Almost all leading jurisdictions apply AI systems in attempts to enhance the efficiency of the court proceedings. Without any doubts, AI already and successfully can imitate activities traditionally performed by humans in the courts: from vision, recognising and extracting information, whether from the document, picture or natural speech, to analysing of information received and predicting the outcomes or decision-making. However, it is hard to say that AI-era in the judiciary has already begun. There is no jurisdiction in the world in which AI is fully given ‘green light’- they are all at the beginning of the AI-journey. That is mostly due to the fact that the same technical specifications, which power achievements, accuracy and flexibility of AI, place serious limitations for the wide application thereof. First of all, AI systems rely on data, which can be biased or spoiled in another way initially or easily manipulated later. Secondly, AI systems are not transparent (black-box-problem) and, as a result, are incomprehensible. These two shortcomings place an obstacle for the correct realisation of some fundamental rights in civil proceedings in their traditional understanding, and consequently, for the wide deployment of AI systems therein. It is concluded that the application of AI in the judiciary, in general, and in the civil proceedings, in particular, is subject of sufficient limitations mostly due to incompliance of AI systems with the traditional understanding of fundamental rights and principles the civil proceedings stand on. In the pursuit of the effectiveness of judiciary by means of AI application, fundamental guarantees can appear at stake, and vice versa, in the pursuit of respect of fundamental rights, the judiciary may be left out of the modern world in the stage of complete inadequacy to the needs of the society, therefore, the issue is required extensive research in order to find a fair and right balance.
Ključne besede: artificial intelligence, judiciary, civil proceedings, AI-judge, efficiency of the judiciary, automatic decision-making
Objavljeno: 24.09.2021; Ogledov: 93; Prenosov: 23
Celotno besedilo (1,13 MB)

Opis: In this research, machine learning algorithms were compared in a landslide-susceptibility assessment. Given the input set of GIS layers for the Starča Basin, which included geological, hydrogeological, morphometric, and environmental data, a classification task was performed to classify the grid cells to: (i) landslide and non-landslide cases, (ii) different landslide types (dormant and abandoned, stabilized and suspended, reactivated). After finding the optimal parameters, C4.5 decision trees and Support Vector Machines were compared using kappa statistics. The obtained results showed that classifiers were able to distinguish between the different landslide types better than between the landslide and non-landslide instances. In addition, the Support Vector Machines classifier performed slightly better than the C4.5 in all the experiments. Promising results were achieved when classifying the grid cells into different landslide types using 20% of all the available landslide data for the model creation, reaching kappa values of about 0.65 for both algorithms.
Ključne besede: landslides, support vector machines, decision trees classifier, Starča Basin
Objavljeno: 13.06.2018; Ogledov: 494; Prenosov: 45
Celotno besedilo (382,76 KB)
Gradivo ima več datotek! Več...

Opis: This paper investigates the relationship between the dimensions of the decision-making process quality and company performance of top managers’ in Slovenia. We found out that companies whose managers exhibit an above-average dimension of openness of spirit in the quality of the decision making process, on average, have a higher stance on foreign markets as companies in which managers show a below-average open spirit. For the managers who work in companies that are present in foreign markets, we could confirm that there is a low/weak correlation between the dimension of effort of the decision-making process quality and the number of employees in a company.
Ključne besede: quality of the decision-making process, rationality, motivation, participation, exhaustivity of the information, managers effort, creativity and innovativeness, company stakeholders, company performance
Objavljeno: 03.05.2018; Ogledov: 699; Prenosov: 95
Celotno besedilo (1,27 MB)
Gradivo ima več datotek! Več...

Opis: This paper describes data structures and algorithms for representation of Boolean functions with reduced ordered binary decision diagrams (ROBDDs). A hash table is used for quick search. Additional information about variables and functions is stored in binary trees. Manipulations on functions are based on a recursive algorithm of ITE operation. The primary goal of this article is describe programming technics needed to realize the idea. For the first time here recursive algorithms for composing functions and garbage collection with a formulae counter are presented. This is better than garbage collection in other known implementations. The results of the tests show that the described representation is very efficient in applications which operate with Boolean functions.
Ključne besede: Reduced Ordered Binary Decision Diagram, Boolean function, logic functions, logic design verification, hash table
Objavljeno: 02.02.2018; Ogledov: 1544; Prenosov: 33
Celotno besedilo (323,32 KB)

Opis: Basic ethical questions, dilemmas and especially decisions do not only affect the life of an individual but can also affect lives of others. In some professional ethics, where decisions about a person's life or death are made, decisions can even be irreversible. In this contribution three ways of deciding by highlighting critical, and reflective decision-making or systematic thought process as the most effective method in ethics have been pointed out. Therefore, taking ethics as a critically reflective morality highlights the fact that we can talk about ethical exploration, so ethics is a process of thinking, not a set of established answers that need only to be passively accepted. It could be concluded that the study of and practice in, evaluating arguments and evidence (moral decision making) via critical thinking as well as using other important skills (raising questions according to Blooms taxonomy and doing a lot of case studies) is the best way to achieve the most fundamental goal in teaching an ethical course -- becoming a better person. And is therefore something that should be in every curriculum.
Ključne besede: authority, critical thinking, ethics, teaching ethics, decision making, intuitions
Objavljeno: 19.12.2017; Ogledov: 435; Prenosov: 62
Celotno besedilo (547,37 KB)
Gradivo ima več datotek! Več...