NOESIS project will conduct a Big Data (“Learning from Data”) exercise in order to understand the factors (the so-called “features”) which are relevant and crucial for generating value from Big Data applications in transport. To realize the concept described and address its objectives, NOESIS methodology is based on three pillars and consisted of the following steps:
Pillar 1: Transport Domain Features and Use cases
- Linking big data products to Transport challenges and use cases.
- Identification of relevant features for generating value from big data investments.
- Set-up of the NOESIS Big Data in Transport Library.
Pillar 2: Learning Framework/Process (Big Data approach of NOESIS)
- Defining the value ranges of features and of the “labels”.
- Control point: Depending on the data available and the definition of values‘ ranges, the appropriate big data techniques will be utilized for classifying the various use cases and identifying patterns.
- Using a combination of machine learning techniques to identify underlying patterns in the data
Pillar 3: Value Capture mechanism
- Definition of a set of evaluation criteria (input to BDTL – Task 2.4).
- Impact Assessment methodology (IAM) (update of BDTL).
- Translating the benefits of IAM into viable Business models in respect to the involved parties.