| Abstract [eng] |
The aim of this master’s thesis is to develop a model for predicting the outcome of financial services disputes using advanced data analytics and artificial intelligence techniques. The main objective is to combine the analysis of legal processes with modern machine learning technologies to increase the accuracy of the prediction and to reduce the costs of legal processes. The thesis includes a review of scientific literature, which examines existing dispute resolution methodologies, the challenges of legal practice and the opportunities for the use of legal technology (LegalTech). A great deal of attention was paid to deep learning models, particularly the BERT algorithm, which is distinguished by its ability to analyze textual data and recognize complex semantic structures. The study used data from public court decision databases, which were carefully processed and prepared for the application of machine learning algorithms. The developed model was evaluated in terms of its accuracy in predicting the outcome of disputes, considering the type of case, the legal arguments of the parties and other relevant factors. The results showed that the BERT model achieved an accuracy of 81.7%, indicating high efficiency in predicting the outcome of financial disputes. This result is competitive in an international context when compared to the results of studies carried out in other countries. The model developed can be applied not only to the analysis of financial disputes, but also to other areas of law, such as labor law or family law, where an accurate prediction of the outcome of disputes would help to streamline the course of legal proceedings. The practical application of the model would allow legal professionals to quickly assess the likely outcome of a dispute, reducing the length and cost of proceedings. The results of the study show that the integration of deep learning technologies into the legal sector is a promising direction that can significantly improve the efficiency of legal processes. |
