Search and Discovery in Legal Document Networks.

Legal practitioners are permanently in the need to search legal collections in order to assess regulatory texts applicable to a specific case. Needs analysis with legal experts [1] showed that users express their needs in the form of complex queries that address both the semantic content and intertextual links between documents. We propose a graph-based search technique that we implement using two different approaches, a structured and a direct one. These approaches are tested on three collections of documents with different size and granularity of document description and a set of queries along with their relevant answers built in collaboration with legal experts.