An idea for Networks and the Law
Posted by shannonclark on May 6, 2005
An idea, some notes and thoughts on it, will work on this further and post more later.
If you take a law firm and do the following, I think there is a possibility of interesting applications of network technology.
First, start by researching the normal patterns of citations in legal cases (Lexis/Nexis has done this in conjunction with an unpublished paper – link will follow)
Using this data, look at the patterns of citation, specifically co-citation (along with results of cases where applicable) and look for the general trends determining whether citations are age related (i.e. frequently cited articles get more citations because of age) vs. the portion of the citations which are due to other causes – i.e. not age but something else such as being a good case and precedent.
Then within a law firm do a similar analysis of the legal documents prepared by the firm (not contracts etc but legal documents which involve citations of precedence – i.e. filings with various courts). Look at the patterns of citation used by the firm – look at the results of those filings (probably available somewhere within the firm’s systems). Compare citation patterns internal to the firm with the general patterns from the Lexis/Nexis data.
My thesis is that when you do this a few things will show up:
– certain patterns of citations which are effective – i.e. core cases which should be cited in a given type of case
– some patterns in citation internal to the firm which are different than the more broadly found patterns in the Lexis/Nexis datasets. Possibly including an ability to determine which pattern is more effective (i.e. it is possible that for some types of cases the patterns of citation used within the firm are more effective than the general pattern seen in the Lexis/Nexis datasets)
This historical analysis could be a good baseline for the firm to then do the following:
A lawyer takes a draft document and sends it to the system.
The system looks at the pattern of citation and determines how the draft document compares with other prior internal documents as well as historical documents from the public record – it can then suggest citations which might logically be included, as well as perhaps citations which might best be removed (i.e. correlate with ineffective filings perhaps).
Obviously this it non-trivial and would only work for the largest of law firms, but I think it might be a fairly interesting and probably valuable system to build – and a possible source of competitive advantage for a firm which used it.