Yanlei Diao is awarded an ERC Consolidator Grant for her "Big Fast Data" project
Yanlei Diao, professor at École Polytechnique and Télécom ParisTech, and member of Inria Saclay – Île-de-France research centre’s team Cedar, has been honoured with a 2016 ERC Consolidator Grant. This European grant rewards the scientific excellence of the researcher and her project, as well as her innovative vision in terms of "acceleration and optimisation of analytical computing for big data". It is the first distinction to concern a researcher who holds an excellence chair funded by the Université Paris-Saclay.
Yanlei Diao, who obtained a PhD from the University of California, Berkeley in 2005, became a professor at the University of Massachusetts Amherst in the United States where she developed research and teaching in the field of big data analytics. Internationally recognised for her expertise in the processing of big data flows, Yanlei Diao also developed numerous collaborations with companies such as Google, IBM, and Cisco. On 1 September 2015, Yanlei Diao was recruited jointly by École Polytechnique and Télécom ParisTech in order to develop a world-class research team in the field of Big Data. Recipient of a University Paris-Saclay excellence chair, Yanlei Diao is a member of the LIX, École Polytechnique's Computer Science Laboratory (under the joint supervision of the CNRS), the team DBWeb of the Télécom ParisTech laboratory LTCI and the team Cedar of Inria Saclay – Île-de-France research centre.
The ERC Consolidator Grant rewards the visionary, ground-breaking research proposed by Yanlei Diao. In the field of Big Data, of paramount importance today is perpectual, low-latency analysis of large volumes of data, referred to as « Big and Fast Data Analysis » in this grant. Such analysis carries the promise to return timely information and insights to numerous applications and to foster revolution in critical scientific (genomics, for example) and societal fields. However, today’s systems cannot support both the scale of data and the low-latency of data analysis. The first set of big data systems (e.g., Hadoop) succeeded in scaling but suffered from high latency of data analysis. The second generation of systems (e.g., Spark) have evolved to cope with live data streams, but they are still struggling with guaranteed latency when the data size exceeds the memory size of a computer cluster. Diao’s research marks the development of a third wave of big data systems that can achieve both scale and low latency for a wide range of complex analytics.
Toward this vision, Diao’s research develops a new algorithmic foundation for clean slate design of big and fast data systems. It advances the state of the art on many fronts. First and foremost, it maximizes the potential of parallelism in all its forms, including both data parallelism (to scale out) and "pipeline" parallelism to reduce processing response times. Furthermore, it designs new algorithmic solutions to enable complex analytical problems, including temporal pattern analysis, graph analysis, and machine learning tasks, to exploit such parallelism to the maximum extent.
("pipeline" refers to a specific process for the parallelisation of calculations, in series)
Diao received training implemented by the Université Paris-Saclay - the support for which she was particularly grateful - when submitting her application for the ERC. Thanks to this funding, Diao will get the chance to surround herself with talented people in order to developed this ambitious project. She also intends to strengthen her collaborations, particularly in Machine Learning.
Diao foresees applications in numerous fields and is seeking real case studies and partner companies in France; she intends to develop collaborations in data management on subjects such as large-scale web data analytics, low-latency genome and healthcare data analysis, real-time object tracking and detection detection in the Internet of Things.
Learn more: http://www.lix.polytechnique.fr/~yanlei.diao/
The ERC Consolidator Grant honours a researcher with 7 to 12 years of post-PhD experience, for the scientific excellence of the researcher and his/her project (value, innovative nature, feasibility, innovative methodology). In total, 314 projects were selected in 2016, including 43 projects based in France.