In this discussion paper we present QA3, a question answering (QA) system over RDF cubes. The system first tags chunks of text with elements of the knowledge base (KB), and then leverages the well-defined structure of data cubes to create the SPARQL query from the tags. For each class of questions with the same structure a SPARQL template is defined. The correct template is chosen by using a set of regex-like patterns, based on both syntactical and semantic features of the tokens extracted from the question. Preliminary results are encouraging and suggest a number of improvements. Over the 50 questions of the QALD-6 challenge, QA3 can process 44 questions, with 0.59 precision and and 0.62 recall, remarkably improving the state of the art in natural language question answering over data cubes.
Querying RDF Data Cubes through Natural Language
Atzori, MaurizioPrimo
;
2018-01-01
Abstract
In this discussion paper we present QA3, a question answering (QA) system over RDF cubes. The system first tags chunks of text with elements of the knowledge base (KB), and then leverages the well-defined structure of data cubes to create the SPARQL query from the tags. For each class of questions with the same structure a SPARQL template is defined. The correct template is chosen by using a set of regex-like patterns, based on both syntactical and semantic features of the tokens extracted from the question. Preliminary results are encouraging and suggest a number of improvements. Over the 50 questions of the QALD-6 challenge, QA3 can process 44 questions, with 0.59 precision and and 0.62 recall, remarkably improving the state of the art in natural language question answering over data cubes.File | Dimensione | Formato | |
---|---|---|---|
sebd18-Querying RDF Data Cubes through Natural Language.pdf
Solo gestori archivio
Tipologia:
versione editoriale (VoR)
Dimensione
238.31 kB
Formato
Adobe PDF
|
238.31 kB | Adobe PDF | Visualizza/Apri Richiedi una copia |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.