Temporal outlier analysis of online civil trial cases based on graph and process mining techniques

Abstract

Since the complete digitisation of civil processes that took place in Italy in 2008, a lot of data regarding the life cycle of thousands of civil proceedings has been collected. However, despite the continuous monitoring that Italian courts are subjected to, and regulatory changes to the procedures that have been enacted in recent years, the average duration of proceedings is still far from acceptable. In order to identify elements which could point to the causes of such 'slowness', data provided by the Court of Livorno have been analysed through process mining and graph techniques, in order to assess the coherence and correct application of the process model. A methodology to identify and analyse the 'outlier' processes has been developed, as described in this work, to also detect characteristics which could justify delays in the processes' completion. In this work, process mining and statistical techniques have been applied to the analysis of the proceedings' data and outliers, with their characteristics, have been recognised.