I was interested to read Dr. Philipp Behrendt's views on how to use IT to manage disputes efficiently - especially how AI may help avoid disputes in the future.
He also refers to the use of AI within document review and how predictive coding utilises algorithms to deliver better and faster results, than a linear human review. We are seeing the analytical techniques that are used to enable predictive coding used on virtually every case we work on. They are becoming an essential tool in the toolkit when tackling a document review exercise of any size and can be used in a multitude of ways from prioritisation to quality control to full-blown predictive coding.
These technologies can also now be used, on the right case, without going through a dedicated learning phase through the use of Continuous Active Learning (CAL). This identifies and prioritises documents likely to be relevant to the matter in hand based on the decisions as the review proceeds. The queue of documents to be reviewed is then re-prioritised throughout the process as the system learns more about what is, and is not, relevant and adjusts accordingly.
The technologies continue to evolve they will continue to improve and be more regularly used in more cases in more ways.
Document review, is the obvious field and we see already that computer assisted document review – predictive coding – delivers better results than human review. Until today, predictive coding is mostly feasible in large cases where a significant number of documents will need to be reviewed as the computer still needs to be trained by a human. The more intelligent the software becomes the lesser training it will need.