An innovative, scientific positioning focused on Data Science

One of Keyrus's strategic priorities is Data Science, the science of processing massive quantities of multi-structured data. Big Data requires that comprehensive approaches be designed for collecting, processing, analysing, categorizing, storing and valorizing data. The methods used for transforming data into structured, integrated information, then into knowledge, are at the heart of this priority. This is how “Data Science” is commonly defined.

With Data Science, companies can enhance and give real, operational value to the use of their Big Data.

Data Science to meet the challenge of Big Data

Data Science is multidisciplinary. It consists of:

  • Applied mathematics: in particular, statistics, probability, linear algebra, etc.
  • Advanced algorithmic-type computer science: Machine Learning, MapReduce, Hadoop, Spark, etc.
  • Management: decision theory, data visualization, etc.

These three disciplines are complementary and indissociable. They form a convergent, complex whole, which can be defined as the science of massive quantities of multi-structured data or Data Science.

Actual uses of Data Science

The actual application of Data Science contributes, among other things, to:

  • The exploiting of metadata to extract its significance and its direction (semantic analysis).
  • Data visualization
  • Large-scale learning of models such as visual models, based on large masses of on-line data.

However, any uncertainty or imprecision in the Data must be integrated in the various stages of managing this data: approaches by intervals, probability, etc. The large-scale, continuous integration of heterogeneous data, with diverse semantics, also requires new techniques of automatic classification. Finally, social relationships between users must be able to be put to good use to improve search quality, in particular with the techniques of recommendations, social graphing, etc.

Data science to meet what challenge?

For businesses, the challenge consists of processing and activating the available data in order to improve their competitiveness. In addition to "standard", so-called "structured" data, which is already handled and exploited by all types of companies through Business Intelligence, there are now massive quantities of unstructured data, coming from the Internet, social media, mobiles and Smart devices. This is a "tsunami" of data or a "dataclysm": more data was produced over just ten years than was ever produced by humanity since its origins.

Keyrus: Associating BI and Big Data Analytics

Keyrus's scientific strategy can be defined as the intention to investigate this new interdisciplinary field of "Data Science", or the Science of Massive, Multi-structured Data. This process lies at the meeting-point of applied mathematics, advanced computer science and management.

This scientific research specialisation is part of the continuing work on Business Intelligence computing in which Keyrus has been a recognized specialist for nearly 20 years.

This strategic positioning guides the Group's R&D policy towards the combined contribution of its original profession, Business Intelligence, and the advent of the analysis of massive, multi-structured data, as much at the methodological level as at the technological one (platform and services).

  • Innovation through data
  • Big Data Architectures and TCO optimization
  • Data Science consulting
  • Big Data Agile Laboratory
  • Big Data Service Factory