An automatic key discovery approach for data linking

Abstract

In the context of Linked Data, different kinds of semantic links can be established between data. However when data sources are huge, detecting such links manually is not feasible. One of the most important types of links, the identity link, expresses that different identifiers refer to the same real world entity. Some automatic data linking approaches use keys to infer identity links, nevertheless this kind of knowledge is rarely available. In this work we propose KD2R, an approach which allows the automatic discovery of composite keys in RDF data sources that may conform to different schemas. We only consider data sources for which the Unique Name Assumption is fulfilled. The obtained keys are correct with respect to the RDF data sources in which they are discovered. The proposed algorithm is scalable since it allows the key discovery without having to scan all the data. KD2R has been tested on real datasets of the international contest OAEI 2010 and on data sets available on the web of data, and has obtained promising results.

SAKey: Scalable Almost Key Discovery in RDF Data

“…In Semantic Web settings where data can be incomplete and may contain multi-valued properties, KD2R [12] aims to derive exact composite keys from a set of non keys discovered on RDF data sources. KD2R, extends [15] to be able to exploit ontologies and consider incomplete data and multivalued properties.…”

“…KD2R, extends [15] to be able to exploit ontologies and consider incomplete data and multivalued properties. Nevertheless, KD2R [12] that is able to discover composite OWL2 keys can be overwhelmed by large datasets and requires clean data. In [3], the authors have developed an approach based on TANE [8] algorithm to discover pseudo-keys (approximate keys) for which a set of few instances may have the same values for the properties of a key.…”

“…In [3], the authors have developed an approach based on TANE [8] algorithm to discover pseudo-keys (approximate keys) for which a set of few instances may have the same values for the properties of a key. The two approaches [12] and [8] differ on the semantics of the discovered keys in case of identity link computation. Indeed, the first considers the OWL2 [13] semantics, where in the case of multivalued properties, to infer an identity link between two instances, it suffices that these instances share at least one value for each property involved in the key, while in [3], the two instances have to share all the values for each property involved in the key, i.e., local completeness is assumed for all the properties (see [2] for a detailed comparison).…”

“…Key discovery approaches have been proposed recently in the setting of the semantic web [3,12]. [3] discovers pseudo keys, keys that do not follow the OWL2 semantics [13] of a key.…”

“…This type of keys appear to be useful when a local completeness of data is known. [12] discovers OWL2 keys in clean data, when no errors or duplicates exist. However, this approach cannot handle the huge amount of data found on the web.…”