The core purpose of the Semantic Web is to ease the search and the navigation of the huge amount of contents currently available on-line in order to reduce the information overload often experimented by Web users. This purpose is achieved by enriching the current Web with semantic descriptions of contents: they are standard-compliant representations of Web data that explicit their meanings in a way suitable to be automatically processed and aggregated. As a consequence software agents, driven by the informative needs of Web users, are enabled to support them in contents browsing by searching, interpreting, integrating and refining data from multiple on-line sources.
In order to realize this vision, it is fundamental to support the creation of huge amounts of semantic data over the Web by exposing online semantic descriptions of contents: these descriptions are built around the possibility to unambiguously identify each concept or entity over the Web by a shared URI. The adoption of the same URI to point out the same concept through distinct online datasets is an important enabling factor of online semantic data integration.
If we consider both search engines like Google or collaborative tagging systems like delicious, currently keywords represents one of the preferred means to describe and search for Web resources: they can be manually or automatically derived from Web data. But keywords are often ambiguous since they can point out many different meanings, thus reducing the recall and precision of Web searches.
To overcome this problem by realizing the possibility to semantically describe and structure information, we can characterize Web resources by one or more concepts instead of simple and ambiguous keywords. To determine the set of concepts describing specific Web contents, we need a properly structured knowledge resource useful to support the disambiguation of the meaning of keywords by pointing out the Semantic Web URI of the intended concepts. Starting from the explanation of why a shared concept URI referencing systems is fundamental in the context of the Semantic Web, we present Tagpedia a semantic reference useful to create semantic descriptions of Web resources: we describe its structure and the way it has been automatically populated by mining Wikipedia. We also briefly introduce Tagpedia online interface and Web API. We point out how Tagpedia exploits the widely adopted DBpedia URI schema so as to identify each concept in the creation of semantic descriptions of Web resources. We also present a relevant application example of Tagpedia. In conclusion, we outline future developments