We are currently rolling out a new data standard for Legal Services groups and courts, on schema.org mark up. This should improve their placement and jurisdiction correctness on searches. We have run a large-scale audit, to document the current problems on Google search, when it comes to people seeking legal help and getting misinformation. We will do a second audit, after the mark-up has been applied, to see how this affect the algorithms placement of legal help resources.
We are also developing a taxonomy of legal problems that people experience in the US. This is being used to build natural language classifiers to automatically spot people's legal issues, in their posts or in their searches. This is in collaboration with Suffolk University's lit lab and Pew charitable trusts.
Read more about the Building a Smarter Legal Internet Search Project