Data Economy unfolding - Part V: Creating Value

The value is generated through data analytics and meta-data customisation that, broadly speaking correspond to the "manufacturing" and "distribution" in the world of atoms.

A variety of companies, many few years old, are populating the space of data leverage, creating value out of Big Data. In this chart companies in the European marketspace. Credit: DataLandscape. EU

Filtering is a crucial part of customisation and an important value generator. Credit: WSJ

The Open Data Framework, ODF, is a powerful backstage that can potentially open up the leveraging of data to third parties. 

Municipalities and Governments are already buying into the Open Data Framework. This is the case for several municipalities in Italy and 'round the world. The IEEE Smart Cities Initiative requires as prerequisite the adoption of the ODF.

Most important, the adoption of the ODF attracts more companies to open their data and this creates local clusters that, although nowhere near the size of the ones of GAFA, are significant enough to be leveraged.

The sheer volume of data available has changed what since the 90ies used to be the technologies to extract information from data. At that time "data mining" was based on statistical analyses of patterns (particularly for fraud detection) along with OLAP (On Line Application Processing) coupled with data base technologies and visualisation technologies.

Today each of the technologies applied to data mining made significant steps forward (visualisation is and will be a fundamental component in value "appreciation"), and we are now relying much more on machine learning, with deep learning technology that is leveraging on neural networks, both simulated and based on chips like Synaptic. In the coming decade we might (no certainty here) benefit from quantum computation to further increase the capability of big data processing.

The extraction of meaning, that is often the result of correlation among different kinds of data originating from different sources, can compare to the manufacturing in the world and economy of atoms. By processing raw materials with ever more sophisticated tools the industrial revolution created value, delivering products that were not possible before, at an affordable price and in volume. By processing huge data clusters using sophisticated hardware and software technologies, the data economy revolution is creating value by delivering information that was not available before, at an affordable price and, basically, with no volume constrain.

There are, however, two differences between manufacturing and data mining.

First, the manufacturing process is somewhat constrained by its own processing tools and related economics to deliver, basically, "one size fits them all". The software tools used for data mining are much more flexible and can perform the mining starting from the client's needs and interest.  It is the customisation of the mining that actually provides valuable information.

This is why when looking at the data economy we can find the value production covering both the "manufacturing" and the "supply-delivery chains", to which we can make the correspondence with "meta-data" production and "customisation". Notice that it makes sense to keep the meta-data production and the customisation separated, even though the meta-data production (analytics) is most of the time driven by the customisation needs. In the chart you can see the number of European data economy companies active in the manufacturing and customisation.

An important aspect of customisation is filtering. Google is a master in this area but there is so much that can be done in filtering, now and even more so in the future as data continues to grow, that there is plenty of space for new players.

Filtering is evolving by embedding awareness and knowledge about ourselves. Interestingly filtering can be performed in stages/layers, each one managed potentially by different players in different parts of the web space. And one most interesting part of the web space is our digital self embedded in our devices.

Second, the manufacturing process "consumes" the raw material (even though we are striving more and more to look at the whole product life cycle with the aim or reusing waste...). This is not so with data. Actually, the creation of information out of data, as result of data mining/deep learning, leads to the availability of even more data (raw material). It is what is sometimes referred to as the "supply-side miracle". The more you use your raw material the more raw material becomes available! This, in a way, is breaking some fundamental economic tenets. We are moving from the economy of scarcity to the economy of abundance.

Author - Roberto Saracco

© 2010-2018 EIT Digital IVZW. All rights reserved. Legal notice