MapMerge: correlating independent schema mappings

Abstract

One of the main steps toward integration or exchange of data is to design the mappings that describe the (often complex) relationships between the source schemas or formats and the desired target schema. In this paper, we introduce a new operator, called MapMerge, that can be used to correlate multiple, independently designed schema mappings of smaller scope into larger schema mappings. This allows a more modular construction of complex mappings from various types of smaller mappings such as schema correspondences produced by a schema matcher or pre-existing mappings that were designed by either a human user or via mapping tools. In particular, the new operator also enables a new "divide-and-merge" paradigm for mapping creation, where the design is divided (on purpose) into smaller components that are easier to create and understand and where MapMerge is used to automatically generate a meaningful overall mapping. We describe our MapMerge algorithm and demonstrate the feasibility of our implementation on several real and synthetic mapping scenarios. In our experiments, we make use of a novel similarity measure between two database instances with different schemas that quantifies the preservation of data associations. We show experimentally that MapMerge improves the quality of the schema mappings, by significantly increasing the similarity between the input source instance and the generated target instance. Finally, we provide a new algorithm that combines MapMerge with schema mapping composition to correlate flows of schema mappings.

The iBench integration metadata generator

“…Example 1.2 (Ad Hoc Scenarios). MapMerge [1] is a system for correlating mappings using overlapping sets of target relations, or using relations that are associated via target constraints. To evaluate their approach, the authors used three real biological schemas (Gene Ontology, UniProt and BioWarehouse) and mapped the first two schemas to the third (BioWarehouse).…”

“…• We demonstrate the power of iBench by presenting a novel evaluation of MapMerge [1], comparing it to two other systems Clio [10] and ++Spicy [18] (Section 7). Our evaluation systematically varies the degree of source and target sharing among mappings in the generated scenarios.…”

“…For mapping creation or adaptation, we can measure the similarity between instances produced by data exchange. Alexe et al [1] proposed a metric for measuring the preservation of data correlations in a target instance wrt. a source instance.…”

Incrementally improving dataspaces based on user feedback

“…Also related to our work, is the MapMerge operator developed by Alexe et al [5]. Given a set of mappings, which are expressed as second order tuple generating dependencies [21], between one (or more) source schema and a target schema, MapMerge correlates those mappings in a meaningful manner.…”

Privacy Preserving Data Integration Through L-Diversity

“…In EIRENE [10], data examples are used to refine schema mappings. The main difficulty when attempting to characterize data transformation using data examples is likely to describe the same behavior.The existing uncorrelated mappings that may result in duplication of data and loss of associations in data exchange.MapMerge [11] exploits constraints in the source and target schemas to find the associations and improves the quality of mappings and increases the scalabilty.In generalization relation null values may occur if it is realized through materializing all specific classes inside a single table and which leads to ambiguous mappings and incorrect data exchange.When exchanging incomplete data and mapping inversion in the source may arise null values [14].The problem of entity fragmentation, and the inability to resolve ambiguous data exchange scenarios caused by different implementations of a generalization relation in source and target, are consequences of ignoring data level mappings. The gap between data level and schema level mappings in schema mapping-based data exchange results in semantic heterogeneities, and consequently, incorrect and redundant target instances.…”