A relational approach for information retrieval on XML legal sources

Abstract

In the information retrieval (IR) domain a collection of documents is represented as a set of documents where cross references between documents are usually not taken into account in the querying process. This standard document model is not tailored to legal professional uses where the context of interpretation is crucial. Existing access tools do not take into account the complexity of references between legal documents. XML based standards have been created to facilitate access and management of legal data. Their exploitation for IR purposes offers new possibilities for advanced search techniques. In this work, we propose a novel approach allowing to exploit the XML standard formats of legal documents to query a collection of related documents and return relevant answers to the end-user. We consider exploiting at the same time the semantic content of the documents and their interrelationships using Formal and Relational Concept Analysis. Answers are presented as documents or graphs of interlinked documents.

Relational Concept Analysis in Practice: Capitalizing on Data Modeling Using Design Patterns

“…Agroecology, and more precisely finding plant-based extracts that can serve as alternative to synthetic pesticides and antimicrobials, was considered in [24]. Using RCA in Information Retrieval (IR) is discussed in [8], and was, for instance, used for querying collections of legal documents connected through cross references [25]. In the field of Semantic Web, it was for instance used to design ontologies [32,20] and extract consistent families of dependent link key candidates in RDF data [4].…”

Introducing LUIMA

“…The system in [21] retrieved documents based on queries containing semantic descriptors and indicators of cross-referential relations between documents (e.g., "Which orders talk about abnormally annoying noise and make reference to decrees talking about soundproofing ? ").…”