Keyword-based search over (semi)structured data is today considered an essential feature of modern information management systems and has become an hot topic in database research and development. Most of the recent approaches to this problem refer to a general scenario where: (i) the data source is represented as a graph, (ii) answers to queries are sub-graphs of the source containing keywords from queries, and (iii) solutions are ranked according to a relevance criteria. In this paper, we illustrate a novel approach to keyword search over semantic data that combines a solution building algorithm and a ranking technique to generate the best results in the first answers generated. We show that our approach is monotonic and has a linear computational complexity, greatly reducing the complexity of the overall process. Finally, experiments demonstrate that our approach exhibits very good efficiency and effectiveness, especially with respect to competing approaches.
DE VIRGILIO, R., Maccioni, A., Cappellari, P. (2013). A Linear and Monotonic Strategy to Keyword Search over RDF Data. In Lecture Notes in Computer Science. BERLIN HEIDELBERG : Springer-Verlag [10.1007/978-3-642-39200-9].
A Linear and Monotonic Strategy to Keyword Search over RDF Data
DE VIRGILIO, ROBERTO;MACCIONI, ANTONIO;
2013-01-01
Abstract
Keyword-based search over (semi)structured data is today considered an essential feature of modern information management systems and has become an hot topic in database research and development. Most of the recent approaches to this problem refer to a general scenario where: (i) the data source is represented as a graph, (ii) answers to queries are sub-graphs of the source containing keywords from queries, and (iii) solutions are ranked according to a relevance criteria. In this paper, we illustrate a novel approach to keyword search over semantic data that combines a solution building algorithm and a ranking technique to generate the best results in the first answers generated. We show that our approach is monotonic and has a linear computational complexity, greatly reducing the complexity of the overall process. Finally, experiments demonstrate that our approach exhibits very good efficiency and effectiveness, especially with respect to competing approaches.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
