Understanding the Semantics of Keyword. Queries on Relational Data Without Accessing the Instance

The simplicity of keyword queries has made them particularly attractive to the technically unskilled user base, tending to become the de-facto standard for querying on the web. Unfortunatelly, alongside its simplicity, came also the loose semantics. Researchers have for a long time studied ways to understand the keyword query semantics and retrieve the most relevant data artifacts. For the web, these artifacts were documents, thus, any semantics discovering effort was based mainly on statistics about the appearance of the keywords in the documents. Recently, there has been an increasing interest in publishing structural data on the web, allowing users to exploit valuable resources that have so far been kept private within companies and organizations. These sources support only structural queries. If they are to become available on the web and be queried, the queries will be in the form of keywords and they will have to be translated into structured queries in order to be executed. Existing works have exploited the instance data in order to build offline an index that is used at query time to drive the translation. This idea is not always possible to implement since the owner of the data source is typically not willing to allow unrestricted access to the data, or to offer resources for the index construction. This chapter elaborates on methods of discovering the semantics of keyword queries without requiring access to the instance data. It describes methods that exploit meta information about the source data and the query in order to find semantic matches between the keywords and the database structures. These matches form the basis for translating the keyword query into a structure query.