SimDataMapper: An Architectural Pattern to Integrate Declarative Similarity Matching into Database Applications

PlanEngine: ferramenta para gerenciamento automatizado de projetos de software

Natália Cristina SchneiderLeonardo Andrade RibeiroAndrei de Souza InácioHarley Michel WagnerAldo vonWangenheim

Effective manipulation of string data is of fundamental importance tomodern database applications. Very often, textual inconsistencies render equalitycomparisons meaningless and strings have to be matched in terms of theirsimilarity. Previous work has proposed techniques to express similarity operationsusing declarative SQL statements. However, the non-trivial issue of embeddingsimilarity support into object-oriented applications has received littleattention. Particularly, a number of indexes have to be maintained for eachsimilarity predicate, thereby severely complicating persistence of applicationobjects. In this paper, we present SimDataMapper, an architectural pattern toprovide easy, efficient, and flexible integration of declarative similarity matchingwith applications and programming environments. We describe implementationdetails and experimentally evaluate the performance of our approach.

Caso o link acima esteja inválido, faça uma busca pelo texto completo na Web: Buscar na Web

Biblioteca Digital Brasileira de Computação - Contato:
     Mantida por: