Charge Prediction by Constitutive Elements Matching of Crimes

Authors: Jie Zhao, Ziyu Guan, Cai Xu, Wei Zhao, Enze Chen

IJCAI 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experiments on two real-world datasets show the superiority of CECP over competitive baselines. We evaluate the performance of CECP on Criminal [Hu et al., 2018] and CAIL [Xiao et al., 2018]
Researcher Affiliation Academia Jie Zhao , Ziyu Guan , Cai Xu , Wei Zhao and Enze Chen School of Computer Science and Technology, Xidian University {jzhao1992@stu., zyguan@, cxu@, ywzhao@mail., 21031211569@stu.}xidian.edu.cn
Pseudocode Yes We alternately optimize them, and the details and the pseudocode of the training process are described in Appendix A.2.
Open Source Code Yes 1Available at: https://github.com/jiezhao6/CECP
Open Datasets Yes We evaluate the performance of CECP on Criminal [Hu et al., 2018] and CAIL [Xiao et al., 2018], which are both collected from the China Judgments Online3.
Dataset Splits Yes Statistics Training Cases Test Cases Charges Criminal-S 61,589 (24.4) 7,702 (24.1) 149 Criminal-M 153,521 (24.4) 19,189 (24.4) 149 Criminal-L 306,900 (24.4) 38,368 (24.5) 149 CAIL 101,275 (22.3) 26,661 (22.2) 119
Hardware Specification No No specific hardware details (GPU/CPU models, memory amounts, or detailed computer specifications) are provided in the paper.
Software Dependencies No No specific ancillary software details (e.g., library or solver names with version numbers) are mentioned in the paper.
Experiment Setup Yes We alternately optimize them, and the details and the pseudocode of the training process are described in Appendix A.2. For CEs construction and more experimental settings, we describe them in Appendix A.3.