rules in artificial intelligence
TRANSCRIPT
© 2015 IBM Corporation
Rules in Artificial Intelligence
Dec 2015 – Presentation at Ecole 42
Pierre Feillet – IBM Decision automation [email protected]
© 2015 IBM Corporation
Agenda
2
Origins
Expert System -> Inference Engine -> Rules
Current state
From raw inference engine to Entreprise decision automation
Business Rules in Bluemix
© 2015 IBM Corporation
RULES TO MIMIC HUMAN MINDFrom Expert Systems to Operation Decision Management
© 2015 IBM Corporation4
Expert Systems
An expert system is a computer system that emulates the decision-making ability of a human expert. Expert systems are designed to solve complex problems by reasoning about knowledge, represented primarily as if–then rules rather than through conventional procedural code.
The first expert systems were created in the 1970s and then proliferated in the 1980s. Expert systems were among the first truly successful forms of AI software.
Expert systems were introduced by the Stanford Heuristic Programming Project. Applied to domains where expertise was highly valued and complex, such as diagnosing infectious diseases (Mycin).
The typical expert system consisted of a knowledge base and an inference engine. The knowledge base stored facts about the world. The inference engine applied logical rules to the knowledge base and deduced
new knowledge. This process would iterate as each new fact in the knowledge base could trigger additional rules in the inference engine.
© 2015 IBM Corporation5
Rule Logic
2 primarily modes of rule inference: forward chaining and backward chaining. Forward chaining starts with the known facts and asserts new facts.
Ex: Socrate is Human so he is mortal Backward chaining starts with goals, and works backward to determine what facts
must be asserted so that the goals can be achieved.
Ex: Is Socrate mortal? It would search through the knowledge base and determine if Socrates was Human and if so would assert he is also Mortal. Can include a common technique was to integrate the inference engine with a
user interface to ask questions when facts are not enough and would then use that information accordingly.
© 2015 IBM Corporation6
Rule Logic
An inference engine cycles through three sequential steps: match rules, select rules, and execute rules. The execution of the rules will often result in new facts or goals being added to the knowledge base which will trigger the cycle to repeat. This cycle continues until no new rules can be matched.
In the first step, match rules, the inference engine finds all of the rules that are triggered by the current contents of the knowledge base. In forward chaining the engine looks for rules where the antecedent (left hand side) matches some fact in the knowledge base. In backward chaining the engine looks for antecedents that can satisfy one of the current goals.
In the second step select rules, the inference engine prioritizes the various rules that were matched to determine the order to execute them.
In the final step, execute rules, the engine executes each matched rule in the order determined in step two and then iterates back to step one again. The cycle continues until no new rules are matched.
Rule engine algorithms: RETE, IBM Fastpath & Sequential, etc
Stateless & stateful processing
© 2015 IBM Corporation7
From Expert Systems to Operational Decision Management
Goal: Empower Business Users to author, test, simulate, deploy their decision logic
Bring a Business Model on the top of Java, XML, JSON, COBOL, etc
Add high level rule artifacts: Decision Table & Trees
Provide near natural language DSLs with editors to write your rules in your preferred locale: Chinese, English
Integrate the rule engine into a server to scale, and hot deploy ruleset in a 24/7 manner
Trace decisions for auditability
Cloud
PaaS & SaaS
Rule engine
BusinessModel
LocalizedBusiness
Languages
Decision warehouse
DecisionServer
Testing & Simulation
BusinessRules
Tools Cloud
© 2015 IBM Corporation
IBM BUSINESS RULESBusiness rules as a service in IBM Bluemix
© 2015 IBM Corporation
Your Application
Externalize Decisions from Applications into Business Rules
Manage decision logic independently from applications
Your Application
Decision logic
Natural language rules can be easily read
Externalized rules are easy to change
Centralized rules enable reuse and consistency
Rules written in software code cannot be read easily
Hard coded rules are difficult to change
Rules intertwined within applications cannot be reused by other systems
Business Rules
© 2015 IBM Corporation
IBM Business Rules, a Smarter Process high value service
Familiar Environment for AuthoringDevelopers can download an Eclipse based authoring tool and author rules in a familiar user-friendly environment.
Separate Business LogicBusiness logic is authored separately from the application which enables easier change in business policy / logic and codified capture of business policies, practices and regulations..
Business logic is easily expressed with business rules to automate decisions with the fidelity of a subject matter expert.
Bridge Business Users and Developers
Deploy Versioned Business LogicMultiple versions of the Business logic can be tested and deployed in the same Business Rules Service. Switching, upgrading, sharing business logic across applications has never been easier.
Enables developers to spend less time recoding and testing when the business policy changes. The Business Rules service minimizes your code changes by keeping business logic separate from application logic.
Business Rules
© 2015 IBM Corporation
The Business Rules service simplifies the experience of creating and managing mobile app business logic – making apps more adaptable
© 2015 IBM Corporation
Developing and deploying applications using the Business Rules service
IBM Bluemix
One app
Another app
Business Rules service instance
Author business rules with Rule Designer plug-ins for Eclipse
Deploy business rules
Develop and push app code
Call the service
Users access apps from their devices
Non-Bluemix apps can call the service too
Call the service
© 2015 IBM Corporation13
Wrap up
Rules
From 70s IA to today enterprise decision management
A large number of companies leverage some kinds of business rules (finance, banking)
Empower developers and business users to automate decision making
Provides transparency and explanation
Dynamic deployment
Rules are based on causality while Big Data & Machine Learning are based on correlation
Perspectives to bridge rules & ML
Try Business Rules in Bluemix https://console.ng.bluemix.net on London or Sydney datacenters
Business Rules
© 2015 IBM Corporation
Q & A
14