OntoJudy: A Ontology Approach for Content-Based Judicial Recommendation Using Particle Swarm Optimisation and Structural Topic Modelling
N. Roopak, Gerard Deepak · Lecture notes in networks and systems · 2021
Under the history of the Judicial Reform of India, major data of judicial cases are commonly used to address the issue of judicial study. Similarity review in legal trials is the foundation of wisdom judicature. The analysis of the Indian Judicial cases in a established format is a central concern by giving necessity for eradicating incompetent information extracting appropriate rules and conditions from the vivid document. Hence, this paper proposes a method to recommend judicial cases to the user, based on the content of the cases. The proposed OntoJudy model uses Static Judicial Domain Ontology with Structural Topic Modelling. The semantic similarity is computed using Particle Swarm Optimization with Jaccard Similarity and SemantoSim. Hybridizing all these helps in yielding better accuracy. To assimilate users preferences for the content recommendations, the CAIL2018 dataset is used which is then classified using Random Forest Classification with the help of extracted query word from the user information. The proposed model has achieved an Accuracy of 95.89% and tends to do better than the other baseline models by attenuating the resilience of the traditional content recommender systems.