Computer Predicts Trial Outcomes with 79 Percent Accuracy
Are we getting closer to using computers instead of judges to decide case outcomes?
A recent study seems to suggest as much.
In the study, researchers fed data on 584 actual cases that had been brought before the European Court of Human Rights into a machine-learning algorithm. The information included topics, circumstances and factual background on each case.
Based on this data, the computer model was able to predict the actual results of the cases with 79 percent accuracy.
Interestingly, the strongest predictive variables were the facts of the case. Legal arguments of counsel were far less helpful.
Improving Court Administration
Observers say the results – in addition to showing how far artificial intelligence has advanced – could help prioritize cases before bodies like the Human Rights Court and identify which matters are most likely to pose serious violations.
“We expect this sort of tool would improve efficiencies of high level, in demand courts,” said one of the scientists.
“We don’t see artificial intelligence replacing judges or lawyers,” said University College London computer scientist Nikolaos Aletras in this press release. Alestras collaborated with researchers from the University of Sheffield and the University of Pennsylvania.
Read the study, published in PeerJ Computer Science, here.
Sources:
- PeerJ Computer Science https://peerj.com/articles/cs-93/
- ABA Journal http://www.abajournal.com/news/article/artificial_intelligence_predicted_case_outcomes_with_79_percent_accuracy_by?utm_source=feedburner&utm_medium=feed&utm_campaign=ABA+Journal+Top+Stories#When:12:00:00Z
University College London https://www.ucl.ac.uk/news/news-articles/1016/241016-AI-predicts-outcomes-human-rights-trials
