Analysis of the effectiveness of the Polish judicial system with the use of topic modeling tools

The rese­arch looks at how to measu­re the effec­ti­ve­ness of the Polish judi­cial sys­tem. This phe­no­me­non was asses­sed using an ana­ly­sis of the like­li­ho­od of appe­aling court deci­sions. The arti­fi­cial intel­li­gen­ce tools used in the stu­dy made it possi­ble to gro­up a lar­ge num­ber of court deci­sions by sub­ject mat­ter. In the gro­ups obta­ined, it was chec­ked how many judg­ments were appe­aled. It also ana­ly­zed what per­cen­ta­ge of judg­ments in a given gro­up were issu­ed by court depart­ments with a given spe­cia­li­za­tion. In this way, gro­ups with homo­ge­ne­ous legal cha­rac­te­ri­stics were iden­ti­fied, i.e. con­cer­ning indi­vi­du­al issu­es of civil, cri­mi­nal, fami­ly or labor law. In addi­tion, such gro­ups of rulings that address more than one legal pro­blem were also obse­rved. Final­ly, it was exa­mi­ned how hete­ro­ge­ne­ous case topics affect the like­li­ho­od of appe­als. The results sug­gest that civil cases are less like­ly to be appe­aled than cri­mi­nal cases. In con­trast, judg­ments with hete­ro­ge­ne­ous sub­ject mat­ter have an extre­me­ly high like­li­ho­od of appe­al. In turn, a bet­ter under­stan­ding of the legal pro­blems asso­cia­ted with the­ma­ti­cal­ly hete­ro­ge­ne­ous gro­ups of judg­ments can serve to impro­ve the ope­ra­tion of the court system.