Publication details
Síťová analýza sémantické podobnosti textů soudních rozhodnutí
| Title in English | Network analysis of semantic similarity of judicial decisions texts |
|---|---|
| Authors | |
| Year of publication | 2019 |
| Type | Appeared in Conference without Proceedings |
| MU Faculty or unit | |
| Citation | |
| Attached files | |
| Description | The topic of this paper is a feasibility study in which semantic similarity methods are experimentally applied to the texts of a transparently selected sample of 200 Supreme Court decisions and the results are processed with the network analysis. Using semantic similarity methods, the texts are measured against each other in order to determine their degree of similarity. Word2vec and cosine similarity methods are applied. From the texts in the selected sample and from the variables that define their mutual similarity, a network is created by network analysis. It is verified whether this network replicates the structure of the thematic interconnection of judicial decisions and their thematic similarity through the clustering analysis. In the clustering analysis, the cluster of labour law decisions is recognized. Finally, the analysis is supplemented by an assessment that the cluster contains all of the labour law decisions in the dataset. However, this method is unable to recognize any other thematic clusters in the network. |