|Anatomy of a GreEd Rule|
|Future Plans (this page)|
We have definite plans for making GreEd useful for diverse uses. Some of the ideas we have, include:
- It will be possible to add new custom operators and functions to the operator palette, which could be more domain specific.
- It will be possible to add custom translator component to allow GreEd to work with a completely new rule languages and engines
- Use of semantically annotated standard data elements to enhance semantic interoperability
- More rule management capabilities like ability to search for rules and navigate the contents of rule repository ‘semantically’. To find a rule of a particular type, you will not need to wade through text of the rules, but will be able to search based on concepts. You could also dynamically alter the organization of the collection of rules so that you can navigate the repository in a way that suits you most.
Examples of dynamic organization based on semantics:
- Grouped by clinical subject areas like Neurology, Ophthalmology, Gastroenterology etc.
- Grouped by clinical categories, e.g., symptoms, signs, lab results etc.
Examples of semantic search :
- Find rules that conclude presence of diabetes
- Find rules that determine patient eligibility = true
- Find rules that use weight loss as an antecedent
- Find rules that utilize ‘disease = papillary serous adenocarcinoma of ovary’ and ‘duration of disease’ as facts
- Find rules that utilize eye conditions (returns rules that use diabetic retinopathy, cataract, corneal ulcer etc.)
- Intelligent automation in creation of rule – GreEd will create a rule set for you, given a set of data elements.
- Intelligent application development – given a set of rules, the application will be able to generate the most optimal set of questions to allow a user to reach an inference
- A software development kit (SDK) to allow you to integrate the GreEd system with your own system so that you can start using its inferencing capabilities in your applications.
- Tools to allow development of intelligent modules using local collection of data elements to provide inferencing for different applications. This will serve as a mechanism for allowing the GreEd system to understand data elements in a local system to allow it to use them for rules creation.