The WISE technology is controlled by target requests (queries defined by the user) that serve as triggers to any processing within the system. When a user is interested in certain data products, it will be first checked if these products have already been processed within the information system. If so, the WISE technology uses techniques to find out whether these data products have been derived using the latest versions of processing algorithms and raw data. If that is the case, the requested targets are delivered to the user. If, on the other hand, the data products were not available, or the raw data/processing algorithms used to generate them was outdated, processing tasks will be started with the new raw data and most updated algorithms. This is what we call "backward chaining".
Backward chaining is made possible, because the WISE technology ensures extreme data lineage by creating and keeping track of all links between each data product and its dependencies back to the raw data. These links are stored via references in a database. Conventional data handling typically involves work flow systems, where information is stored about the algorithms used for the generation of data products and the associated flow of events. Such methods fail when arbitrary software programs that reside outside the data handling system,are used for data processing. WISE technology tackles the problem by making the database, where all the references about data products are stored an integral part of all data processing. Such close integration guarantees the high datalineage (including both the raw data and all metadata associated with every data product), which is crucial to the success of the WISE technology. Backward chaining also brings the flexibility to introduce new software to process existing data, without the need to update or rebuild the existing software environment.
|Last modified:||02 October 2015 10.56 p.m.|