Legal search is broken – can it be fixed with open source taxonomies?

I spent yesterday afternoon at the International Society for Knowledge Organisation‘s Legal KnowHow event, a series of talks on legal knowledge and how it is managed. The audience was a mixture of lawyers, legal information managers, vendors and academics, and the talks came from those who are planning legal knowledge systems or implementing them. I also particularly enjoyed hearing from Adam Wyner from Liverpool University who is modelling legal arguments in software, using open source text analysis. You can see some of the key points I picked up on our Twitter feed.

What became clear to me during the afternoon is that search technology is not currently serving the needs of lawyers or law firms. The users want a simple Google-like interface (or think they do), the software is having trouble presenting results in context and the source data is large, complex and unwieldy. The software used for search is from some of the biggest commercial search vendors (legal firms seem to ‘follow the pack’ in terms of what vendor they select – unfortunately few of the large law firms seem to have even considered the credible open source alternatives such as Lucene/Solr or Xapian).

In many cases taxonomies were presented as the solution – make sure every document fits tidily into a heirarchy and all the search problems go away, as lawyers can simply navigate to what they need. All very simple in theory – however each big law firm and each big legal information publisher has their own idea of what this taxonomy should be.

After the final presentation I argued that this seemed to be a classic case where an open source model could help. If a firm, or publisher were prepared to create an open source legal taxonomy (and to be fair, we’re only talking about 5000 entries or so – this wouldn’t be a very big structure) and let this be developed and improved collaboratively, they would themselves benefit from others’ experience, the transfer of legal data between repositories would be easier and even the search vendors might learn a little about how lawyers actually want to search. The original creators would be seen as thought-leaders and could even license the taxonomy so it could not be rebadged and passed off as original by another firm or publisher.

However my plea fell on stony ground: law firms seem to think that their own taxonomies have inherent value (and thus should never be let outside the company) and they regard the open source model with suspicion. Perhaps legal search will remain broken for the time being.

Share this postShare on FacebookShare on Google+Tweet about this on TwitterShare on LinkedInShare on RedditEmail this to someone

2 thoughts on “Legal search is broken – can it be fixed with open source taxonomies?

  1. I agree 100%, besides for the standard industry codes that we try to incorporate in most of our DMS solutions the rest of the taxonomy remains completely unique for each law firm. I don’t know why this is, you would think that they would all agree on a standard taxonomy and just use that. We have done a few mergers and getting the taxonomies the same is a nightmare.

Leave a Reply

Your email address will not be published. Required fields are marked *