An essential language for declarative business rules. Lex Wedemeijer Open University in the Netherlands

Size: px

Start display at page:

Download "An essential language for declarative business rules. Lex Wedemeijer Open University in the Netherlands"

Eleanor Hill
8 years ago
Views:

1 An essential language for declarative business rules Lex Wedemeijer pen University in the Netherlands

2 An essential language for declarative business rules Lex Wedemeijer pen University in the Netherlands

3 Agenda Business Rules Rule-engineering approach Language specification: steps Characteristics Discussion

4 Business Rules describe business operations in 'natural' way many types of rules ECA-rules, triggers, workflow-rules, derivation rules, transition rules, pre- and post conditions,... declarative rules: state-oriented, time-invariant novice learners

derivation rules, transition rules, pre- and post conditions,.

5 Business Rules business context design artifacts computersupported operations languages: natural business jargon controlled / semi formal exact specifications computer lingo

6 Example: at the IT helpdesk natural business jargon: every call should get an acceptable response controlled language: keyword concept for every call placed_by a client there MUST be a response that is available_for the call which is accepted_by the client exact specs: (Binary Relation Algebra) fact-type computer lingo: (ruleml; SWRL; PRR)

client there MUST be a response that is available_for the call which is accepted_by

7 Binary Relation Algebra sound math suited to declarative rules non-computer lingo main components cardinality rules compound rules Rules rule violations concepts s Data Model atoms tuples recorded business data

cardinality rules compound rules Rules rule

8 Business Rules business context design artifacts computersupported operations languages: natural business jargon controlled / semi formal exact specifications computer lingo

9 Rule-engineering approach business context Rules Repository design artifacts Process specs Datamodel specs computersupported operations Application Software Design Generation Business procedures Application Generation workflow operations Business Service definitions Interface definitions recorded business data

Software Design Generation Business procedures Application Generation

10 Rule-engineering approach business context Rules Repository Data Model workflow operations recorded business data define Data Model specify the rules test rules with data test overall process deliver to developers

business data define Data Model specify the rules

11 Rule-engineering approach define language has 5 statements rule populate enforce explain define the Model specify the rules test rules with data test overall process deliver to developers

12 Specification: step 1 concept define define concept -name concept (infix notation) Unconstrained Conceptual Model Structure = structural constraints Business Vocabulary

13 Specification: step 2 concept define define rule cardinality_rule compound_rule

14 Specification: step 2 detail cardinality _rule rule rule-id must be keyword simple_constraint and

15 Specification: step 2 detail cardinality _rule rule rule-id must be simple_constraint and unary-operator expression expression binary-operator expression ( expression ) compound _rule rule rule-id for internal reference expression must comparator expression keyword imply, be implied compound s

binary-operator expression ( expression ) compound _rule rule rule-id for

16 Specification: step 3 concept define define rule cardinality_rule compound_rule populate populate concept { { atom tuple } } Entity integrity, Referential Integrity Rule violations emerge

17 Specification: step 4 concept define define rule cardinality_rule compound_rule populate populate concept { { atom tuple } } enforce enforce rule-id immediate deferred Rule violations: prevent or permit not: how to amend

concept { { atom tuple } } enforce enforce rule-id

18 Specification: any step concept define define rule cardinality_rule compound_rule populate populate concept { { atom tuple } } enforce enforce rule-id immediate deferred concept explain explain quoted_text rule-id

19 Characteristics 1. essential (orthogonal) 2. step-by-step 3. notations kept simple 4. compares to RuleSpeak (almost understandable)

20 Example: at the IT helpdesk define [call] placed_by [client], [response] accepted_by [client], [response] available_for [call] rule 1_cardinal [call] placed_by [client] must be univalent and total rule 4_helpdesk [call] placed_by [client] must imply [call] available_for~ [response];[response] accepted_by [client] populate [call] placed_by [client] { thursday#1 * lex, fri#2 * kim } populate [response] available_for [call] { reply#77 * thursday#1 } enforce 1_cardinal immediate enforce 4_helpdesk deferred explain 4_helpdesk "IF a call is placed_by a client THEN must a response be available for that call AND that response must be accepted_by the client"

$[call] placed_by [client] { thursday#1 * lex, fri#2 * kim } populate [response] available_for [call] { reply#77 * thursday#1 } enforce 1_cardinal immediate enforce$

21 Characteristics: first findings 9 novice students, no prior experience in rule engineering compare "IT helpdesk" scripts in rich language (Ampersand, see versus essential language

22 First findings rich language versus essential language is more suitable for new rule designers may be easier to learn for new rule designers is more educational (learning while designing) is easier to use when creating a new rule-based design is better aligned to stepwise design approach may cause more confusion compared with component-oriented language:.. is easier to explain to co-workers novice students (N=9) is simpler to check for errors is better in avoiding conflicting (inconsistent) statements? has more brief and powerful statements

23 Agenda Business Rules Rule-engineering approach Language specification: steps Characteristics Discussion

24 Thank you IF every question has got an acceptable-answer THEN the speaker thanks you for your attention

Content chapter 2. Ampersand

Content chapter 2. Ampersand Content chapter 2 1 Introduction 21 2 Rules 21 3 Rules in Business 23 4 An example 24 5 Control Principle 26 6 Rule Management 27 7 Case Study: A Service Desk 27 7.1 The business rules 28 7.2 Specification