Recommendations in an online shop such as “other customers also bought…” is an example of a data product we are nowadays coming across with great regularity. Another example of a data product is a fraud detection system; this can trace dishonest reviews on a comparing website. With the help of a data product, users can make faster, more efficient, more effective and more accurate decisions. In the applied research centre Data Science practical-based research is done into the creation of data products.
Data products for the benefit of delta areas
Disciplines such as mathematics, statistics, software engineering and machine learning are joined in research. This is combined with knowledge about a specific domain. HZ’s applied research centre Data Science focuses on the domain knowledge in the region south western Delta. It aims at making data products, with which insight is created in the delta system. Think for example of protection and safety, food, tourism, industry, energy and logistics. For this reason, the applied research centre Data Science collaborates with businesses, local governments and other applied research centres active in these domains.
The research process to create a data product
Every data product is the result of a research process. The first crucial step, business understanding, aims at investigating what exactly is the question. After that, rough data are collected and adapted to the further process. A possible next step is visualising, or in another way communicating, the data. Then, the data can be further modelled by means of machine learning. This entails various techniques such as clustering, predicting, classifying, detecting anomalies, etc. Ultimately, this will result in a data product.