Rule ML 2017 - 12/07/17 London
Interesting talks here in London @ Rule ML, We started with the keynote about Deriving Rules from Data by Elena Baralis. She described different techniques to generalise concepts on top of a dataset to be able to extract rules from these data sets. It was quite interesting to see how these patterns and techniques were used to analyse any type of transactional datasets. These examples included a really nice use case for Documents Summarisation, which used Weighted and Generalised Association Rules.
[gallery ids="3329,3330,3331,3332,3333" type="rectangular"]
Then during the morning I attended a really nice session about Semantic DMN, which was about how to integrate DMN decision tables with Domain Knowledge based on ontologies and more advanced concepts. These guys created a bridge which combines the information that is represented by domain models to columns defined in the decision tables. This augmentation provided a lot of interesting and additional features to what you would expect from a typical DMN implementation. I’m pretty sure that this is were DMN should go in subsequent versions.
Just before lunch Time Aware Business Processes was all about a semantic model to check time consuming Tasks in BPMN processes. The semantic model demonstrated provided a way to check how much time you will need to perform several tasks and to make sure that our processes adhere to these constraints or we get notified otherwise. Unfortunately, this was a very academic project, with a prolog implementation only :(
After lunch, I joined an LPS session from Imperial Collage. LPS stands fro Logic-based production systems a concept that if you are familiar with Production Systems and rule engines you should check.
LPS distinguish between two type of concepts Goals and Beliefs: Goal: if a person is a on a bus then the person has a ticket -> this is a reactive rule that will be looking to hold as a regulation Belief: a person has a ticket if the person is on a bus -> This is related to the ontology, the concepts that we have
For a more detailed description you can check their bitbucket project lpsmasters / lps_corner And I would recommend to check the slides as well because they contain a lot of very interesting history facts: Artificial Intelligence and Human ThinkingAnd check out their live demo HERE