Big data identification and calculation model construction of civil abuse of right of action behavior based on the constraints of laws and regulations

Abstract Civil abuse of rights of action belongs to the violation of the purpose of civil litigation, excessive consumption of judicial resources. It violates the principle of good faith of the matter litigation, not only has a serious impact on the trial authority to carry out the work, and wastes the judicial resources, but also constitutes damage to the interests of other subjects of the litigation, and also adversely affects the stability of the social order. For this reason, this paper analyzes the civil abuse of litigation rights based on legal and regulatory constraints, and constructs a big data identification model using LightGBM algorithm. The study first starts from the balancing of data, and the feature engineering of civil abuse of rights of action is processed to improve the performance of the model. For the binary classification problem of civil abusive litigation behavior, through the training and adjustment of the model, the model accuracy is more than 90%, and the average AUC, ACC, and Recall of the model in the ten ten-fold cross-validation results are 0.9201, 0.9518, and 0.9173, respectively, and the performance is better compared with other models. The identification model constructed not only ensures the accuracy of identification, but also improves the imbalance defects of the data of civil abuse of right of action line, which is more applicable to the situation of civil abuse of right of action in China, so that countermeasures against the status quo of civil abuse of right of action behavior can be proposed based on the identification results.