Repository files navigation

Capstone Proposal for Udacity Machine Learning Nanodegree Program.

When you think of the legal industry, what is the first thing that comes to your mind? Machine Learning? Probably not. When I started working in the industry as a Data Scientist one year ago, I was met with suspicion by lawyer colleagues who were eager to point out reasons why a computer wouldn't be able to do their jobs. It turns out that they were right sometimes, but not all of the time, fortunately.

Why is it fortunate? Aside from keeping my job, those were refreshing news because Brazilian Courts have reached99.7 million pending lawsuits in 2014 and numbers continue to increase. Machine Learning can help companies stay competitive in this scenario of increasing litigation costs by freeing lawyers (and other knowledge workers) from menial, laborous and often error-prone tasks and letting them focus on work that require high-order reasoning.

Beyond automation of simple tasks, Machine Learning can also be used to augment knowledge workers performance by providing complementary context to the task at hand, such as:

• Highligh important information in a document
• Search for documents according to a query
• Recommend similar documents

This processes is usually refered to asTechnology Assisted Review (TAR).