RNTI

MODULAD
Natural language interfaces for data warehouses
In EDA 2012, vol. RNTI-B-8, pp.83-92
Résumé
Data warehouses process a huge mass of data, that are often modeled in a way humans hardly understand. In order to make the search paradigm more accessible to end users, some efforts have been made in the field of Business Intelligence. However, expressing information need as a structured query is still an artificial task. This explains why "natural" approaches are preferred nowadays. In this paper, we present a Question Answering (Q&A) system for structured data in a context of BI. Some benefits of our proposal are described through an iPhone/iPad application and through an HTML prototype.