Predictive Coding Demystified | eDiscovery News

Paralegals with ten or more years experience will remember the days when we sat in conference rooms along side contract lawyers reviewing hundreds of boxes of documents, reading page by page looking for specific information. Later, we begin to sit at computers staring at those pages of documents which had been scanned into a database - still looking for specific information while entering keywords, names, and other such information for each document and summarizing the documents. These two exercises were called "document review" and required a "thinking" person with knowledge about the case to read and comprehend the documents, while defining privilige, issue, and relevance.

Then there was "coding". These are the rooms full of "coders", possessing only basic reading skills, and with little understanding of what they were looking at and why or if it was important. These data entry workers captured objective data by entering key words and dates, addresses and names mentioned. Indeed, they are warned not to waste time reading, but to quickly scan a page, looking for specific words only.  Their performance was measured by number of keystrokes and documents completed per minute.  This was necessary before OCR (optical scanning recognition) became reliable.

Now, there's predictive coding. What is it? Have we all been replaced by artificial intelligence? eDiscovery News describes it best at this link - Predictive Coding Demystified eDiscovery News.

No comments:

Post a Comment