Advanced data analytics: For transparent and responsible AI
With the world becoming more and more connected, data is becoming increasingly important. This has all kinds of consequences and offers a lot of possibilities. TNO has expertise in the field of data sharing and analysis and considers its mission to provide insight and meaning to data in an explainable and responsible way.
The role of data in our society and economy is growing steadily more important. People and devices are generating ever more data. The increase in the number of sensors, smartphones and wearables, for example, provides even more data sources. So there is a growing demand for reliable and insightful data analysis, data governance, data standards and data infrastructures for sharing and analysing data. "We help companies and governments share, manage and make better decisions from both a social and a business perspective in a responsible and transparent way," explains Timon Brussaard, business developer for Advanced Data Analytics.
The Advanced Data Analytics roadmap deals with all kinds of big data processes in which artificial intelligence (AI) is important. Within our three knowledge areas - Explainable, Responsible and Interoperable Data Science - the most important priority is always to give insight and meaning to data in an understandable and responsible way. For example, we focus on making AI-based data analysis methods transparent. We develop automatic reasoning technologies that can derive relationships in data and explain the outcome of the analysis to people. Because data may contain sensitive information, we develop solutions with a privacy-friendly design. In addition, we realise solutions for the efficient and controlled sharing of digital information. People are at the heart of all this.
Within the Curium project, data from juvenile psychiatric clients are analysed to support the therapist with diagnoses. The emphasis is on making the analysis explainable: how does the model arrive at a statement, and what makes a certain diagnosis more likely than another in this case?
Patient Forum Miner is a text mining application that makes it possible to make information on patient forums searchable and interpretable for medical professionals. Especially for rare diseases, there are often too few patients in one country for good research, and international forums where patients share their experiences can be a good source of information.
Another area in which our knowledge is deployed is the analysis of criminal activities. In the European research projects ASGARD and TITANIUM, TNO is working on the analysis of digital traces of criminal activities, particularly on the dark web, in virtual currencies such as bitcoin, and on confiscated devices. For example, a profile is created on the basis of messages that can say something about the author, so that it is possible to trace messages on different sites to the same person.
If you would like to know more about what we do in the field of Advanced Analytics or would you like to work with us, please read below or contact us.
Big Data Value Center
The Big Data Value Center (shorted: BDVC) is a unique 'open innovation platform' with a physical place in Almere, The Netherlands. In the Big Data Value Center, we create more value out of (big) data....