BI Composition

Coralie Haese -
Multimedia Communications

Date: -
Location: Eurecom

The aim of this work is to study a new way of creating BI analysis by especially targeting casual users. We called this operation BI Composition and it consists of reusing parts of existing BI snippets (charts or tables) to merge them together to obtain a new snippet. This operation is not trivial at all because the accurate merging of content from different sources depends on many factors like for example data compatibility or data summarizability. These challenges were identified during this internship and I will present the first analysis I made on it. To showcase this BI Composition, I implemented a prototype, The Composer, which is a client-server application developed in Flex and Java EE. This proof of concept enabled to assess the potential of BI composition: study the possible gestures and implement the first composition use cases. Further work will be conducted on this topic in the next years to extend the current use cases and to try to tackle the problems of data merging. Bio: Coralie Haese is PhD candidate at EURECOM / SAP Sophia Antipolis. This presentation will consist of showing the work I achieved during my six month internship ending my engineering studies as well as outlining the topics I will work on during my PhD. I joined the Business Intelligence (BI) team at SAP Research Labs in Sophia-Antipolis and drove an innovative project in relation with several teams: the research team, the development and the user experience teams.