Recording of law enforcement needs to measure the extent to what the enforcement law has been carried out in accordance with applicable legal rules. This law enforcement record can be obtained by collecting, analyzing and concluding various legal cases from court decision documents. The number of court decision documents is vast and continues to grow at any time; it requires an extraction process of information that can be run automatically. The conditions of court decision documents that have different use of sentences and the use of word variations are challenges in the process of information extraction. This research proposed rule-based information extraction for court decision documents to exploring and developing the recording of law enforcement in Indonesia. The process uses three steps, identifying structure extraction, tokenization and extracting entities. This research use 150 court decision documents dataset and the experiment has produced 0.82, 0.96, 0.89, beside of average recall, precision, and f-score. This shows that rule-based information extraction can be implemented in the legal domain and supports the development of law enforcement records in Indonesia.